{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.4, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0, "model.model.language_model.layers.26.mlp.down_proj": 1.125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.59375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.1171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.09375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 0 }, { "epoch": 0.0008, "model.model.language_model.layers.26.mlp.down_proj": 1.1328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.59375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.5078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.95703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 1 }, { "epoch": 0.0016, "model.model.language_model.layers.26.mlp.down_proj": 1.109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.5, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.09375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 2 }, { "epoch": 0.0024, "model.model.language_model.layers.26.mlp.down_proj": 1.1328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.09375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.0625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 3 }, { "epoch": 0.0032, "model.model.language_model.layers.26.mlp.down_proj": 1.1171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.03125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.94140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 4 }, { "epoch": 0.004, "model.model.language_model.layers.26.mlp.down_proj": 1.125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.1015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.09375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 5 }, { "epoch": 0.0048, "model.model.language_model.layers.26.mlp.down_proj": 1.1328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.46875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.1015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 6 }, { "epoch": 0.0056, "model.model.language_model.layers.26.mlp.down_proj": 1.109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.95703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 7 }, { "epoch": 0.0064, "model.model.language_model.layers.26.mlp.down_proj": 1.125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 8 }, { "epoch": 0.0072, "model.model.language_model.layers.26.mlp.down_proj": 1.0859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.03125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 9 }, { "epoch": 0.008, "grad_norm": 0.9462440013885498, "learning_rate": 4.986666666666667e-05, "loss": 11.50390625, "step": 10 }, { "epoch": 0.008, "model.model.language_model.layers.26.mlp.down_proj": 1.1953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.09375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.03125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 1.0078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 10 }, { "epoch": 0.0088, "model.model.language_model.layers.26.mlp.down_proj": 1.21875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.5625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.3359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.1171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 1.046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 11 }, { "epoch": 0.0096, "model.model.language_model.layers.26.mlp.down_proj": 1.1015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.09375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.4140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 12 }, { "epoch": 0.0104, "model.model.language_model.layers.26.mlp.down_proj": 1.0859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.0625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.91796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 13 }, { "epoch": 0.0112, "model.model.language_model.layers.26.mlp.down_proj": 1.1875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 2.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 3.15625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 1.0, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 14 }, { "epoch": 0.012, "model.model.language_model.layers.26.mlp.down_proj": 1.0703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.9296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.1484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.9375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.91796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 15 }, { "epoch": 0.0128, "model.model.language_model.layers.26.mlp.down_proj": 1.0625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.90625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.1328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 16 }, { "epoch": 0.0136, "model.model.language_model.layers.26.mlp.down_proj": 1.078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.03125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 17 }, { "epoch": 0.0144, "model.model.language_model.layers.26.mlp.down_proj": 1.0625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.75, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.78125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.9140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 18 }, { "epoch": 0.0152, "model.model.language_model.layers.26.mlp.down_proj": 1.0390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.7265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.0859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.34375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.75, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.90234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 19 }, { "epoch": 0.016, "grad_norm": 1.204301357269287, "learning_rate": 4.973333333333334e-05, "loss": 10.540625, "step": 20 }, { "epoch": 0.016, "model.model.language_model.layers.26.mlp.down_proj": 1.03125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.0703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.90625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 20 }, { "epoch": 0.0168, "model.model.language_model.layers.26.mlp.down_proj": 1.0625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.0546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.03125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.90234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 21 }, { "epoch": 0.0176, "model.model.language_model.layers.26.mlp.down_proj": 1.03125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.5859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.0546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.65625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 22 }, { "epoch": 0.0184, "model.model.language_model.layers.26.mlp.down_proj": 1.03125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.5078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 1.015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 23 }, { "epoch": 0.0192, "model.model.language_model.layers.26.mlp.down_proj": 1.0546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.5390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.65625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 24 }, { "epoch": 0.02, "model.model.language_model.layers.26.mlp.down_proj": 1.0234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.98046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.98828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.65625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 25 }, { "epoch": 0.0208, "model.model.language_model.layers.26.mlp.down_proj": 1.0625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.9453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.9765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.84375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.90234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 26 }, { "epoch": 0.0216, "model.model.language_model.layers.26.mlp.down_proj": 1.015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.96484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.3046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 1.0078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.65625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 27 }, { "epoch": 0.0224, "model.model.language_model.layers.26.mlp.down_proj": 1.015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.92578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.9609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.87890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 28 }, { "epoch": 0.0232, "model.model.language_model.layers.26.mlp.down_proj": 0.97265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.90625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.98046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.53125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 29 }, { "epoch": 0.024, "grad_norm": 1.1400302648544312, "learning_rate": 4.96e-05, "loss": 9.33359375, "step": 30 }, { "epoch": 0.024, "model.model.language_model.layers.26.mlp.down_proj": 0.97265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.88671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.96484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.5, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.87890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 30 }, { "epoch": 0.0248, "model.model.language_model.layers.26.mlp.down_proj": 0.95703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.2109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.8671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.9375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 31 }, { "epoch": 0.0256, "model.model.language_model.layers.26.mlp.down_proj": 0.97265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.8828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.96484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.46875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.87890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 32 }, { "epoch": 0.0264, "model.model.language_model.layers.26.mlp.down_proj": 0.953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.8671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.4375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 33 }, { "epoch": 0.0272, "model.model.language_model.layers.26.mlp.down_proj": 0.94921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.1484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.84765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.9296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.85546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 34 }, { "epoch": 0.028, "model.model.language_model.layers.26.mlp.down_proj": 1.09375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.15625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.94140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.65625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.90625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 35 }, { "epoch": 0.0288, "model.model.language_model.layers.26.mlp.down_proj": 0.9921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.09375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.83984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.95703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 36 }, { "epoch": 0.0296, "model.model.language_model.layers.26.mlp.down_proj": 0.92578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.0703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.92578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 37 }, { "epoch": 0.0304, "model.model.language_model.layers.26.mlp.down_proj": 0.90625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.03125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.7578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 38 }, { "epoch": 0.0312, "model.model.language_model.layers.26.mlp.down_proj": 0.90234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.0546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.78125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.91015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 39 }, { "epoch": 0.032, "grad_norm": 1.0937861204147339, "learning_rate": 4.9466666666666665e-05, "loss": 8.35234375, "step": 40 }, { "epoch": 0.032, "model.model.language_model.layers.26.mlp.down_proj": 0.9140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.76953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 40 }, { "epoch": 0.0328, "model.model.language_model.layers.26.mlp.down_proj": 0.91796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.76953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 41 }, { "epoch": 0.0336, "model.model.language_model.layers.26.mlp.down_proj": 0.93359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.8046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.92578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 42 }, { "epoch": 0.0344, "model.model.language_model.layers.26.mlp.down_proj": 0.875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 1.015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.75390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 43 }, { "epoch": 0.0352, "model.model.language_model.layers.26.mlp.down_proj": 0.8671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.98046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.71484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.86328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.0625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 44 }, { "epoch": 0.036, "model.model.language_model.layers.26.mlp.down_proj": 0.875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.96875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.71875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.0625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 45 }, { "epoch": 0.0368, "model.model.language_model.layers.26.mlp.down_proj": 0.89453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.98828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.9296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 2.0625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 46 }, { "epoch": 0.0376, "model.model.language_model.layers.26.mlp.down_proj": 0.859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.9296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.6953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.1015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.85546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 47 }, { "epoch": 0.0384, "model.model.language_model.layers.26.mlp.down_proj": 0.85546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.90234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.69140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.09375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.9765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 48 }, { "epoch": 0.0392, "model.model.language_model.layers.26.mlp.down_proj": 0.81640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.65625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.9375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 49 }, { "epoch": 0.04, "grad_norm": 1.0280265808105469, "learning_rate": 4.933333333333334e-05, "loss": 7.49140625, "step": 50 }, { "epoch": 0.04, "model.model.language_model.layers.26.mlp.down_proj": 0.83203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.8828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.8515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.8046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 50 }, { "epoch": 0.0408, "model.model.language_model.layers.26.mlp.down_proj": 0.80859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.8359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.62109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.7890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.78125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 51 }, { "epoch": 0.0416, "model.model.language_model.layers.26.mlp.down_proj": 0.8203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.87109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.66015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.81640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.8828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 52 }, { "epoch": 0.0424, "model.model.language_model.layers.26.mlp.down_proj": 0.78125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.8203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.59765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.77734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.78125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 53 }, { "epoch": 0.0432, "model.model.language_model.layers.26.mlp.down_proj": 0.79296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.83984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.64453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 54 }, { "epoch": 0.044, "model.model.language_model.layers.26.mlp.down_proj": 0.79296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.84375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.6484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.80859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.8515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.79296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 55 }, { "epoch": 0.0448, "model.model.language_model.layers.26.mlp.down_proj": 0.8046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.8125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.60546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.99609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.78515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.7421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.78125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 56 }, { "epoch": 0.0456, "model.model.language_model.layers.26.mlp.down_proj": 0.78515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.8125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.63671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.80078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.71875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.77734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 57 }, { "epoch": 0.0464, "model.model.language_model.layers.26.mlp.down_proj": 0.80078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.79296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.6171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.77734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.7265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.76171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 58 }, { "epoch": 0.0472, "model.model.language_model.layers.26.mlp.down_proj": 0.75390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.75, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.58203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.76953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.6640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 59 }, { "epoch": 0.048, "grad_norm": 1.1246705055236816, "learning_rate": 4.92e-05, "loss": 6.701953125, "step": 60 }, { "epoch": 0.048, "model.model.language_model.layers.26.mlp.down_proj": 0.76171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.7734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.60546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.96484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.80859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.5703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 60 }, { "epoch": 0.0488, "model.model.language_model.layers.26.mlp.down_proj": 0.75390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.78125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.61328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.6171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.76171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 61 }, { "epoch": 0.0496, "model.model.language_model.layers.26.mlp.down_proj": 0.76953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.80078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.98046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.75390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.5703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 62 }, { "epoch": 0.0504, "model.model.language_model.layers.26.mlp.down_proj": 0.75390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.76171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.58984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 1.0078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.5703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 63 }, { "epoch": 0.0512, "model.model.language_model.layers.26.mlp.down_proj": 0.82421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.71484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.55078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.94921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.74609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.6875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 64 }, { "epoch": 0.052, "model.model.language_model.layers.26.mlp.down_proj": 0.75390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.75390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.60546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.94921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.74609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 65 }, { "epoch": 0.0528, "model.model.language_model.layers.26.mlp.down_proj": 0.69921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.71875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.53515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.9453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.71484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.46875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 66 }, { "epoch": 0.0536, "model.model.language_model.layers.26.mlp.down_proj": 0.765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.73046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.5703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.91796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.7734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.4296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 67 }, { "epoch": 0.0544, "model.model.language_model.layers.26.mlp.down_proj": 0.7265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.7421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.5625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.9453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.4375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.73046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 68 }, { "epoch": 0.0552, "model.model.language_model.layers.26.mlp.down_proj": 0.67578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.6953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.5234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.90234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.71484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 69 }, { "epoch": 0.056, "grad_norm": 1.215430498123169, "learning_rate": 4.906666666666667e-05, "loss": 6.060546875, "step": 70 }, { "epoch": 0.056, "model.model.language_model.layers.26.mlp.down_proj": 0.6796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.gate_proj": 0.69921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.mlp.up_proj": 0.52734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.k_proj": 0.8984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.o_proj": 0.71484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.q_proj": 1.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 1.0, "model.model.language_model.layers.26.self_attn.v_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 1.0, "step": 70 }, { "epoch": 0.0568, "model.model.language_model.layers.26.mlp.down_proj": 0.7265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.85009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.93310546875, "model.model.language_model.layers.26.mlp.up_proj": 0.5546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.88720703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.75, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.944091796875, "model.model.language_model.layers.26.self_attn.q_proj": 1.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.959716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.876220703125, "step": 71 }, { "epoch": 0.0576, "model.model.language_model.layers.26.mlp.down_proj": 0.6875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.808837890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.91552734375, "model.model.language_model.layers.26.mlp.up_proj": 0.5234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.9140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.880615234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.9296875, "model.model.language_model.layers.26.self_attn.q_proj": 1.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.950439453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.8486328125, "step": 72 }, { "epoch": 0.0584, "model.model.language_model.layers.26.mlp.down_proj": 0.71875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.9033203125, "model.model.language_model.layers.26.mlp.up_proj": 0.55859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.8173828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.864501953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.75, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.911865234375, "model.model.language_model.layers.26.self_attn.q_proj": 1.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.942138671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.71484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.81982421875, "step": 73 }, { "epoch": 0.0592, "model.model.language_model.layers.26.mlp.down_proj": 0.70703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.7626953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.6953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.895751953125, "model.model.language_model.layers.26.mlp.up_proj": 0.53515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.87890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.8544921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.73046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.902099609375, "model.model.language_model.layers.26.self_attn.q_proj": 1.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.93798828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.70703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.803466796875, "step": 74 }, { "epoch": 0.06, "model.model.language_model.layers.26.mlp.down_proj": 0.734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.739990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.75, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.59765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.75390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.8916015625, "model.model.language_model.layers.26.self_attn.q_proj": 1.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.933349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.779052734375, "step": 75 }, { "epoch": 0.0608, "model.model.language_model.layers.26.mlp.down_proj": 0.68359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.68359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.877685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.5234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.764892578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.85546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.88427734375, "model.model.language_model.layers.26.self_attn.q_proj": 1.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.93115234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.6875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.7646484375, "step": 76 }, { "epoch": 0.0616, "model.model.language_model.layers.26.mlp.down_proj": 0.66015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.71728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.873779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.478515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.755126953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.8046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.73828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.876953125, "model.model.language_model.layers.26.self_attn.q_proj": 1.1328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.927978515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.751220703125, "step": 77 }, { "epoch": 0.0624, "model.model.language_model.layers.26.mlp.down_proj": 0.6484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.6484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.49609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.74609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.71875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.86962890625, "model.model.language_model.layers.26.self_attn.q_proj": 1.140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.92431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.6875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.739990234375, "step": 78 }, { "epoch": 0.0632, "model.model.language_model.layers.26.mlp.down_proj": 0.6953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.6640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.mlp.up_proj": 0.515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.73583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.8125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.82275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.q_proj": 1.109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.922119140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.6796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.724853515625, "step": 79 }, { "epoch": 0.064, "grad_norm": 1.2443997859954834, "learning_rate": 4.8933333333333335e-05, "loss": 5.51328125, "step": 80 }, { "epoch": 0.064, "model.model.language_model.layers.26.mlp.down_proj": 0.68359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.mlp.up_proj": 0.5234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.72802734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.8359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.81982421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.71484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.q_proj": 1.1171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.918701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.68359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.713623046875, "step": 80 }, { "epoch": 0.0648, "model.model.language_model.layers.26.mlp.down_proj": 0.6328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.6640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.mlp.up_proj": 0.49609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.72265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.80859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.815673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.q_proj": 1.0625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.91748046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.70166015625, "step": 81 }, { "epoch": 0.0656, "model.model.language_model.layers.26.mlp.down_proj": 0.69921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.66796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.mlp.up_proj": 0.53515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.7158203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.self_attn.o_proj": 0.73828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.q_proj": 1.0703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.91552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.68359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.69140625, "step": 82 }, { "epoch": 0.0664, "model.model.language_model.layers.26.mlp.down_proj": 0.65625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.6875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.mlp.up_proj": 0.5078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.8095703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.q_proj": 1.1015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.9140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.682373046875, "step": 83 }, { "epoch": 0.0672, "model.model.language_model.layers.26.mlp.down_proj": 0.734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.mlp.up_proj": 0.50390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.80078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.805419921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.q_proj": 1.1171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.91162109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.6640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.66357421875, "step": 84 }, { "epoch": 0.068, "model.model.language_model.layers.26.mlp.down_proj": 0.71875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.653076171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.7109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.858154296875, "model.model.language_model.layers.26.mlp.up_proj": 0.5390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.80078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.69140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.q_proj": 1.0390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.910400390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.68359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.652099609375, "step": 85 }, { "epoch": 0.0688, "model.model.language_model.layers.26.mlp.down_proj": 0.6640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.63671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.854248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.46484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.687255859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.7578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.69140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.self_attn.q_proj": 1.0546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.90673828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.626953125, "step": 86 }, { "epoch": 0.0696, "model.model.language_model.layers.26.mlp.down_proj": 0.7109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.633544921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.69921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.854248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.78125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.73046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.self_attn.q_proj": 1.0234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.90576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.6796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.62060546875, "step": 87 }, { "epoch": 0.0704, "model.model.language_model.layers.26.mlp.down_proj": 0.61328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.628662109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.66015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.4921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.7734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.96484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.90478515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.65234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.615478515625, "step": 88 }, { "epoch": 0.0712, "model.model.language_model.layers.26.mlp.down_proj": 0.7421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.616455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.85009765625, "model.model.language_model.layers.26.mlp.up_proj": 0.486328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.674072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.76953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.63671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.812255859375, "model.model.language_model.layers.26.self_attn.q_proj": 1.046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.90380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.66015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.596435546875, "step": 89 }, { "epoch": 0.072, "grad_norm": 1.154853343963623, "learning_rate": 4.88e-05, "loss": 5.038671875, "step": 90 }, { "epoch": 0.072, "model.model.language_model.layers.26.mlp.down_proj": 0.75, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.61328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.73046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.mlp.up_proj": 0.5859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.670166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.78125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.78271484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.809326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.9609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.903076171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.5869140625, "step": 90 }, { "epoch": 0.0728, "model.model.language_model.layers.26.mlp.down_proj": 0.671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.605712890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.mlp.up_proj": 0.46484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.665283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.7421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.77978515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.64453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.902099609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.57275390625, "step": 91 }, { "epoch": 0.0736, "model.model.language_model.layers.26.mlp.down_proj": 0.6328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.60107421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.458984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.798095703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.901611328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.564453125, "step": 92 }, { "epoch": 0.0744, "model.model.language_model.layers.26.mlp.down_proj": 0.67578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.597412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.64453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.mlp.up_proj": 0.5078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.75, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.89453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.901123046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.55517578125, "step": 93 }, { "epoch": 0.0752, "model.model.language_model.layers.26.mlp.down_proj": 0.625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5947265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.mlp.up_proj": 0.470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.75, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.87890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.54931640625, "step": 94 }, { "epoch": 0.076, "model.model.language_model.layers.26.mlp.down_proj": 0.70703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.592041015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.68359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.mlp.up_proj": 0.55859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.653076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.76953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.772705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.7421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.791259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.88671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.544189453125, "step": 95 }, { "epoch": 0.0768, "model.model.language_model.layers.26.mlp.down_proj": 0.671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5888671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.62109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8447265625, "model.model.language_model.layers.26.mlp.up_proj": 0.486328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.6513671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.77001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.85546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.537841796875, "step": 96 }, { "epoch": 0.0776, "model.model.language_model.layers.26.mlp.down_proj": 0.61328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.585693359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8447265625, "model.model.language_model.layers.26.mlp.up_proj": 0.48046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.649169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.76904296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.8359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.900146484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.5322265625, "step": 97 }, { "epoch": 0.0784, "model.model.language_model.layers.26.mlp.down_proj": 0.64453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5830078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.6484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.mlp.up_proj": 0.5078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.646728515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.768798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.78271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.8515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.899658203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.52685546875, "step": 98 }, { "epoch": 0.0792, "model.model.language_model.layers.26.mlp.down_proj": 0.578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.57373046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.5625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.843017578125, "model.model.language_model.layers.26.mlp.up_proj": 0.4375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.639892578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.67578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7646484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.65625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.5087890625, "step": 99 }, { "epoch": 0.08, "grad_norm": 1.0102007389068604, "learning_rate": 4.866666666666667e-05, "loss": 4.71953125, "step": 100 }, { "epoch": 0.08, "model.model.language_model.layers.26.mlp.down_proj": 0.59765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.843017578125, "model.model.language_model.layers.26.mlp.up_proj": 0.453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.638671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.6875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.763671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.501220703125, "step": 100 }, { "epoch": 0.0808, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.56689453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.6171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.843017578125, "model.model.language_model.layers.26.mlp.up_proj": 0.4609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.63623046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.6796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.770263671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.79296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.49658203125, "step": 101 }, { "epoch": 0.0816, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.564208984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.mlp.up_proj": 0.490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.634033203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.68359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.76123046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.768310546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.77734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.62109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.4931640625, "step": 102 }, { "epoch": 0.0824, "model.model.language_model.layers.26.mlp.down_proj": 0.625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.561279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.mlp.up_proj": 0.462890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.6318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.67578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.76025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.766357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.7734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.48828125, "step": 103 }, { "epoch": 0.0832, "model.model.language_model.layers.26.mlp.down_proj": 0.6328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5576171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.4765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.63037109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.66796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.759033203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.76318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.79296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.483154296875, "step": 104 }, { "epoch": 0.084, "model.model.language_model.layers.26.mlp.down_proj": 0.70703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.55224609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.6171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.mlp.up_proj": 0.494140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.6259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.6796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7568359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.8203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.470947265625, "step": 105 }, { "epoch": 0.0848, "model.model.language_model.layers.26.mlp.down_proj": 0.78125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.544677734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.mlp.up_proj": 0.5078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.620849609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.753662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.75439453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.460205078125, "step": 106 }, { "epoch": 0.0856, "model.model.language_model.layers.26.mlp.down_proj": 0.61328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.541259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.62109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83837890625, "model.model.language_model.layers.26.mlp.up_proj": 0.470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.618408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.6640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.752197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.753662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.76171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.45556640625, "step": 107 }, { "epoch": 0.0864, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.535888671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.5390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.mlp.up_proj": 0.431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.6142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.75146484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.75048828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.446533203125, "step": 108 }, { "epoch": 0.0872, "model.model.language_model.layers.26.mlp.down_proj": 0.6484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.533935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.6328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.mlp.up_proj": 0.478515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.613037109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.65234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.75, "model.model.language_model.layers.26.self_attn.o_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.74853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.76953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.44287109375, "step": 109 }, { "epoch": 0.088, "grad_norm": 0.9183698296546936, "learning_rate": 4.853333333333334e-05, "loss": 4.4390625, "step": 110 }, { "epoch": 0.088, "model.model.language_model.layers.26.mlp.down_proj": 0.68359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.53125, "model.model.language_model.layers.26.mlp.gate_proj": 0.64453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.mlp.up_proj": 0.48828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.611572265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.6484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.749267578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.64453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.746826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.76171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.43896484375, "step": 110 }, { "epoch": 0.0888, "model.model.language_model.layers.26.mlp.down_proj": 0.65625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.526123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.53125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.mlp.up_proj": 0.451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.60791015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.741455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.71875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.426513671875, "step": 111 }, { "epoch": 0.0896, "model.model.language_model.layers.26.mlp.down_proj": 0.65625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.mlp.up_proj": 0.5, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.66015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.738525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.4228515625, "step": 112 }, { "epoch": 0.0904, "model.model.language_model.layers.26.mlp.down_proj": 0.66796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.520263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.60546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.mlp.up_proj": 0.47265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.60546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.6171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74658203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.73681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.75, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.58203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.417236328125, "step": 113 }, { "epoch": 0.0912, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.517822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.5703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.mlp.up_proj": 0.466796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.602783203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74560546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.735595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.58203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.4130859375, "step": 114 }, { "epoch": 0.092, "model.model.language_model.layers.26.mlp.down_proj": 0.60546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.516357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.mlp.up_proj": 0.45703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.601806640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74462890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.734619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.41064453125, "step": 115 }, { "epoch": 0.0928, "model.model.language_model.layers.26.mlp.down_proj": 0.61328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.515380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.5, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.601318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.64453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74462890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.7333984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.7421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.407470703125, "step": 116 }, { "epoch": 0.0936, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.51220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.6171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.834716796875, "model.model.language_model.layers.26.mlp.up_proj": 0.4609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.599853515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.6171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74462890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.58984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.73291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.72265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.406005859375, "step": 117 }, { "epoch": 0.0944, "model.model.language_model.layers.26.mlp.down_proj": 0.5859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.51025390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.62109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.834228515625, "model.model.language_model.layers.26.mlp.up_proj": 0.482421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.59765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.743896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.62109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.7314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.4013671875, "step": 118 }, { "epoch": 0.0952, "model.model.language_model.layers.26.mlp.down_proj": 0.54296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.57421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.mlp.up_proj": 0.404296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74267578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.729736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.66796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.399169921875, "step": 119 }, { "epoch": 0.096, "grad_norm": 0.7250761389732361, "learning_rate": 4.8400000000000004e-05, "loss": 4.2076171875, "step": 120 }, { "epoch": 0.096, "model.model.language_model.layers.26.mlp.down_proj": 0.482421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.505126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.53515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.833740234375, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.594970703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.742431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.6484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.898681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.39501953125, "step": 120 }, { "epoch": 0.0968, "model.model.language_model.layers.26.mlp.down_proj": 0.59765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.5859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.4296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.593505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.74169921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.59375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.7265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.67578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.392578125, "step": 121 }, { "epoch": 0.0976, "model.model.language_model.layers.26.mlp.down_proj": 0.59375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.501708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.mlp.up_proj": 0.435546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.592529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.741455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.62109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.72509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.38916015625, "step": 122 }, { "epoch": 0.0984, "model.model.language_model.layers.26.mlp.down_proj": 0.64453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.5, "model.model.language_model.layers.26.mlp.gate_proj": 0.60546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.4921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.592041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.58984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.741455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.69921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.58203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.3876953125, "step": 123 }, { "epoch": 0.0992, "model.model.language_model.layers.26.mlp.down_proj": 0.6171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.493896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.586181640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.740966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.7265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.37890625, "step": 124 }, { "epoch": 0.1, "model.model.language_model.layers.26.mlp.down_proj": 0.640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.492431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.61328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.mlp.up_proj": 0.466796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.584716796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.740966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.718994140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.6953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89794921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.5625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.376708984375, "step": 125 }, { "epoch": 0.1008, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.584716796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.53125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.740478515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.718017578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89794921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.371826171875, "step": 126 }, { "epoch": 0.1016, "model.model.language_model.layers.26.mlp.down_proj": 0.5859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.48779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.60546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73974609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.62109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.716064453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.66796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89794921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.5625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.365966796875, "step": 127 }, { "epoch": 0.1024, "model.model.language_model.layers.26.mlp.down_proj": 0.5859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.58203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.582275390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.71533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.6640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.897705078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.36328125, "step": 128 }, { "epoch": 0.1032, "model.model.language_model.layers.26.mlp.down_proj": 0.62109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.48388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.45703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5810546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.739013671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.63671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.71435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.897705078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.359619140625, "step": 129 }, { "epoch": 0.104, "grad_norm": 0.6238394379615784, "learning_rate": 4.826666666666667e-05, "loss": 4.035546875, "step": 130 }, { "epoch": 0.104, "model.model.language_model.layers.26.mlp.down_proj": 0.71875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.5859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.71240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.7109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8974609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.352783203125, "step": 130 }, { "epoch": 0.1048, "model.model.language_model.layers.26.mlp.down_proj": 0.58203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.477783203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.43359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.577392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.738037109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8974609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.3505859375, "step": 131 }, { "epoch": 0.1056, "model.model.language_model.layers.26.mlp.down_proj": 0.5546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.49609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.574462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73779296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.70654296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.64453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.897216796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.341552734375, "step": 132 }, { "epoch": 0.1064, "model.model.language_model.layers.26.mlp.down_proj": 0.68359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.47119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.62890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.50390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.572998046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.737060546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89697265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.337646484375, "step": 133 }, { "epoch": 0.1072, "model.model.language_model.layers.26.mlp.down_proj": 0.59765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.59765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.mlp.up_proj": 0.470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.572021484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.736572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.62109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.65234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896728515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.334716796875, "step": 134 }, { "epoch": 0.108, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4677734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.mlp.up_proj": 0.447265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.57080078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.736083984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.6015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.700927734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.6328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896728515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.53125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.331298828125, "step": 135 }, { "epoch": 0.1088, "model.model.language_model.layers.26.mlp.down_proj": 0.5703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.466064453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.58203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.mlp.up_proj": 0.42578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.569580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.700439453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896728515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.32861328125, "step": 136 }, { "epoch": 0.1096, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.44921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.56884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.735595703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896728515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.325927734375, "step": 137 }, { "epoch": 0.1104, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.mlp.up_proj": 0.48828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.568359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.735107421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.66015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.32421875, "step": 138 }, { "epoch": 0.1112, "model.model.language_model.layers.26.mlp.down_proj": 0.6953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.6328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.mlp.up_proj": 0.52734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.56640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.734130859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896240234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.59375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.318359375, "step": 139 }, { "epoch": 0.112, "grad_norm": 0.5624966621398926, "learning_rate": 4.8133333333333336e-05, "loss": 3.883984375, "step": 140 }, { "epoch": 0.112, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.58203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.mlp.up_proj": 0.45703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.565673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.734130859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.693603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.64453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.896240234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.53125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.31591796875, "step": 140 }, { "epoch": 0.1128, "model.model.language_model.layers.26.mlp.down_proj": 0.62890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.45654296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.427734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.56298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.734130859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.689208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.66015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89599609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.3095703125, "step": 141 }, { "epoch": 0.1136, "model.model.language_model.layers.26.mlp.down_proj": 0.498046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.5, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.559814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8955078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.5, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.302490234375, "step": 142 }, { "epoch": 0.1144, "model.model.language_model.layers.26.mlp.down_proj": 0.490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.449951171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.55078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.558349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8955078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.301025390625, "step": 143 }, { "epoch": 0.1152, "model.model.language_model.layers.26.mlp.down_proj": 0.625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.44775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.6015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.44921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.557861328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73291015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8955078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.29833984375, "step": 144 }, { "epoch": 0.116, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.55859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.41796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.55712890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73291015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8955078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.29736328125, "step": 145 }, { "epoch": 0.1168, "model.model.language_model.layers.26.mlp.down_proj": 0.515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.49609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.556396484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.732666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.895263671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.29638671875, "step": 146 }, { "epoch": 0.1176, "model.model.language_model.layers.26.mlp.down_proj": 0.6171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.55517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.732177734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.894775390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.2939453125, "step": 147 }, { "epoch": 0.1184, "model.model.language_model.layers.26.mlp.down_proj": 0.5859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.44140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.55078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.553466796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.482421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73193359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.59375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.894775390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.29150390625, "step": 148 }, { "epoch": 0.1192, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.440673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.5703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.447265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.552490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7314453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.289794921875, "step": 149 }, { "epoch": 0.12, "grad_norm": 0.4854952394962311, "learning_rate": 4.8e-05, "loss": 3.7525390625, "step": 150 }, { "epoch": 0.12, "model.model.language_model.layers.26.mlp.down_proj": 0.625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.439697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.5859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.462890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.552001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73095703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.673095703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.62890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.287841796875, "step": 150 }, { "epoch": 0.1208, "model.model.language_model.layers.26.mlp.down_proj": 0.5546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.438232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.54296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.41796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.55126953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.59765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.286376953125, "step": 151 }, { "epoch": 0.1216, "model.model.language_model.layers.26.mlp.down_proj": 0.5703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.41015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5498046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.73046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.58203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.284912109375, "step": 152 }, { "epoch": 0.1224, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.58984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.466796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.549072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.730224609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.28369140625, "step": 153 }, { "epoch": 0.1232, "model.model.language_model.layers.26.mlp.down_proj": 0.578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.53515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.416015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.548583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.729736328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.28076171875, "step": 154 }, { "epoch": 0.124, "model.model.language_model.layers.26.mlp.down_proj": 0.55859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8291015625, "model.model.language_model.layers.26.mlp.up_proj": 0.388671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.546630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7294921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.894287109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.27880859375, "step": 155 }, { "epoch": 0.1248, "model.model.language_model.layers.26.mlp.down_proj": 0.56640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.5078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.828857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.544677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.729248046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.665283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89404296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.27587890625, "step": 156 }, { "epoch": 0.1256, "model.model.language_model.layers.26.mlp.down_proj": 0.58203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.54296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.828857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.543701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.53125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.59375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.2744140625, "step": 157 }, { "epoch": 0.1264, "model.model.language_model.layers.26.mlp.down_proj": 0.54296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.428466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.54296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.419921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.541748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.663330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.58203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.893310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.272705078125, "step": 158 }, { "epoch": 0.1272, "model.model.language_model.layers.26.mlp.down_proj": 0.54296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.53515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.40625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.53955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728271484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.893310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.498046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.27001953125, "step": 159 }, { "epoch": 0.128, "grad_norm": 0.4404528737068176, "learning_rate": 4.7866666666666674e-05, "loss": 3.619140625, "step": 160 }, { "epoch": 0.128, "model.model.language_model.layers.26.mlp.down_proj": 0.59765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.44921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.539306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728271484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.58984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.893310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.267822265625, "step": 160 }, { "epoch": 0.1288, "model.model.language_model.layers.26.mlp.down_proj": 0.62890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.53857421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728271484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.893310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.26611328125, "step": 161 }, { "epoch": 0.1296, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.41015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.537841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.728271484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.893310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.265380859375, "step": 162 }, { "epoch": 0.1304, "model.model.language_model.layers.26.mlp.down_proj": 0.55859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.400390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.537353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72802734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.53125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89306640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.26416015625, "step": 163 }, { "epoch": 0.1312, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.458984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.357421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.533935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.727783203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.65380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89306640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.259765625, "step": 164 }, { "epoch": 0.132, "model.model.language_model.layers.26.mlp.down_proj": 0.5546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.54296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.828369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.533203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.727783203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.653076171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89306640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.258544921875, "step": 165 }, { "epoch": 0.1328, "model.model.language_model.layers.26.mlp.down_proj": 0.56640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.828125, "model.model.language_model.layers.26.mlp.up_proj": 0.419921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5322265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.65185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89306640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.482421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.258056640625, "step": 166 }, { "epoch": 0.1336, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.51171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.53125, "model.model.language_model.layers.26.self_attn.k_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.256591796875, "step": 167 }, { "epoch": 0.1344, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.400390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.531005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.649658203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.254638671875, "step": 168 }, { "epoch": 0.1352, "model.model.language_model.layers.26.mlp.down_proj": 0.56640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.51953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.419921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.530517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.58984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.648681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.253662109375, "step": 169 }, { "epoch": 0.136, "grad_norm": 0.4294019043445587, "learning_rate": 4.773333333333333e-05, "loss": 3.5640625, "step": 170 }, { "epoch": 0.136, "model.model.language_model.layers.26.mlp.down_proj": 0.546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.408203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.64599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.25146484375, "step": 170 }, { "epoch": 0.1368, "model.model.language_model.layers.26.mlp.down_proj": 0.515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.384765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.529541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.64501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.25048828125, "step": 171 }, { "epoch": 0.1376, "model.model.language_model.layers.26.mlp.down_proj": 0.640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.409423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.56640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.4609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52880859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.727294921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.64404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.61328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.249755859375, "step": 172 }, { "epoch": 0.1384, "model.model.language_model.layers.26.mlp.down_proj": 0.5234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.40869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.51953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.384765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.528076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726806640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.45703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.5625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892822265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.247314453125, "step": 173 }, { "epoch": 0.1392, "model.model.language_model.layers.26.mlp.down_proj": 0.49609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.51953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892333984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.246826171875, "step": 174 }, { "epoch": 0.14, "model.model.language_model.layers.26.mlp.down_proj": 0.515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.41796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.892333984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.244873046875, "step": 175 }, { "epoch": 0.1408, "model.model.language_model.layers.26.mlp.down_proj": 0.494140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.470703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52685546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.63916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89208984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.24267578125, "step": 176 }, { "epoch": 0.1416, "model.model.language_model.layers.26.mlp.down_proj": 0.578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.54296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.427734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.526611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89208984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.5, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.2412109375, "step": 177 }, { "epoch": 0.1424, "model.model.language_model.layers.26.mlp.down_proj": 0.515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.63720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.891845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.46875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.239501953125, "step": 178 }, { "epoch": 0.1432, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.400390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.525634765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.57421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.635498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.891845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.2392578125, "step": 179 }, { "epoch": 0.144, "grad_norm": 0.3633517920970917, "learning_rate": 4.76e-05, "loss": 3.46640625, "step": 180 }, { "epoch": 0.144, "model.model.language_model.layers.26.mlp.down_proj": 0.51171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.525146484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.891845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.23779296875, "step": 180 }, { "epoch": 0.1448, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3994140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.53125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.40234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.524658203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.634033203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.23681640625, "step": 181 }, { "epoch": 0.1456, "model.model.language_model.layers.26.mlp.down_proj": 0.5546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.827392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.40234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.726318359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.632568359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.891357421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.235595703125, "step": 182 }, { "epoch": 0.1464, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.51171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826904296875, "model.model.language_model.layers.26.mlp.up_proj": 0.396484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72607421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.632080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.54296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.46875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.233642578125, "step": 183 }, { "epoch": 0.1472, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.466796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82666015625, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.52294921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725830078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.631103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.232421875, "step": 184 }, { "epoch": 0.148, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.390380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.41796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826416015625, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.521240234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.627685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.890380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.228515625, "step": 185 }, { "epoch": 0.1488, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.44921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826416015625, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5205078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.627197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.890380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.228515625, "step": 186 }, { "epoch": 0.1496, "model.model.language_model.layers.26.mlp.down_proj": 0.4921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.5, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826171875, "model.model.language_model.layers.26.mlp.up_proj": 0.390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.520263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.627197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.890380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.22705078125, "step": 187 }, { "epoch": 0.1504, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.386962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.49609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826171875, "model.model.language_model.layers.26.mlp.up_proj": 0.373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.520263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.62646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.890380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.22607421875, "step": 188 }, { "epoch": 0.1512, "model.model.language_model.layers.26.mlp.down_proj": 0.51171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.38623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.486328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.826171875, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.624755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.53125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.890380859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.225830078125, "step": 189 }, { "epoch": 0.152, "grad_norm": 0.3548739552497864, "learning_rate": 4.746666666666667e-05, "loss": 3.3744140625, "step": 190 }, { "epoch": 0.152, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.5234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.396484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.519287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.725341796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.89013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.225341796875, "step": 190 }, { "epoch": 0.1528, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.385009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.55859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.447265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.519287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.58984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.623779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.5703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.889892578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.224365234375, "step": 191 }, { "epoch": 0.1536, "model.model.language_model.layers.26.mlp.down_proj": 0.58203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.435546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51904296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.60546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.2236328125, "step": 192 }, { "epoch": 0.1544, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.49609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.517822265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.62255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.889404296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.223388671875, "step": 193 }, { "epoch": 0.1552, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.46875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.517333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.62109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.22216796875, "step": 194 }, { "epoch": 0.156, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.38134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.517333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.724853515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.620849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.45703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.220458984375, "step": 195 }, { "epoch": 0.1568, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.380615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.474609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.62060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.4375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.21923828125, "step": 196 }, { "epoch": 0.1576, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.5, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.515869140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6201171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.218017578125, "step": 197 }, { "epoch": 0.1584, "model.model.language_model.layers.26.mlp.down_proj": 0.5625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.37939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.53125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.423828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.514404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.724365234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.61962890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.55859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.217529296875, "step": 198 }, { "epoch": 0.1592, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.378173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.53515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72412109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.618896484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.21728515625, "step": 199 }, { "epoch": 0.16, "grad_norm": 0.4026135504245758, "learning_rate": 4.7333333333333336e-05, "loss": 3.3248046875, "step": 200 }, { "epoch": 0.16, "model.model.language_model.layers.26.mlp.down_proj": 0.484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72412109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.216064453125, "step": 200 }, { "epoch": 0.1608, "model.model.language_model.layers.26.mlp.down_proj": 0.609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.53515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.44921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.513427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.723876953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.616455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.215087890625, "step": 201 }, { "epoch": 0.1616, "model.model.language_model.layers.26.mlp.down_proj": 0.48046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3759765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.482421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.616455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.21484375, "step": 202 }, { "epoch": 0.1624, "model.model.language_model.layers.26.mlp.down_proj": 0.515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.512939453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.615478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.53125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.214111328125, "step": 203 }, { "epoch": 0.1632, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.511962890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.723388671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.614013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888916015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.435546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.212646484375, "step": 204 }, { "epoch": 0.164, "model.model.language_model.layers.26.mlp.down_proj": 0.474609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.45703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.337890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.511474609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.723388671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.613525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.211669921875, "step": 205 }, { "epoch": 0.1648, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.511474609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72314453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.5546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.61328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.211181640625, "step": 206 }, { "epoch": 0.1656, "model.model.language_model.layers.26.mlp.down_proj": 0.4921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.3828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5107421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.722900390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.613037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.210693359375, "step": 207 }, { "epoch": 0.1664, "model.model.language_model.layers.26.mlp.down_proj": 0.54296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825927734375, "model.model.language_model.layers.26.mlp.up_proj": 0.408203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.51025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.722900390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.209716796875, "step": 208 }, { "epoch": 0.1672, "model.model.language_model.layers.26.mlp.down_proj": 0.59375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.5078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825439453125, "model.model.language_model.layers.26.mlp.up_proj": 0.40234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.508544921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.721923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.611572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88818359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.20703125, "step": 209 }, { "epoch": 0.168, "grad_norm": 0.3212536573410034, "learning_rate": 4.72e-05, "loss": 3.238671875, "step": 210 }, { "epoch": 0.168, "model.model.language_model.layers.26.mlp.down_proj": 0.52734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.36572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825439453125, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.50830078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.721435546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.611572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.887939453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.206298828125, "step": 210 }, { "epoch": 0.1688, "model.model.language_model.layers.26.mlp.down_proj": 0.490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.44140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.825439453125, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.506103515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.4921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.202880859375, "step": 211 }, { "epoch": 0.1696, "model.model.language_model.layers.26.mlp.down_proj": 0.58203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8251953125, "model.model.language_model.layers.26.mlp.up_proj": 0.423828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.505615234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.60888671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.53515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.201416015625, "step": 212 }, { "epoch": 0.1704, "model.model.language_model.layers.26.mlp.down_proj": 0.45703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.824951171875, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.504150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.720947265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.608154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.20068359375, "step": 213 }, { "epoch": 0.1712, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.462890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.824951171875, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.503662109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.720947265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.607666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.40625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.199951171875, "step": 214 }, { "epoch": 0.172, "model.model.language_model.layers.26.mlp.down_proj": 0.4921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.503662109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.720947265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.60693359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.886962890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.199462890625, "step": 215 }, { "epoch": 0.1728, "model.model.language_model.layers.26.mlp.down_proj": 0.435546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.416015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82421875, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.720458984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.605224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.197265625, "step": 216 }, { "epoch": 0.1736, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82421875, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.720458984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6044921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.19677734375, "step": 217 }, { "epoch": 0.1744, "model.model.language_model.layers.26.mlp.down_proj": 0.5234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.5, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82421875, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.50146484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.60400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.195556640625, "step": 218 }, { "epoch": 0.1752, "model.model.language_model.layers.26.mlp.down_proj": 0.52734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.51171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.5, "model.model.language_model.layers.26.self_attn.k_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.886474609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1943359375, "step": 219 }, { "epoch": 0.176, "grad_norm": 0.2959813177585602, "learning_rate": 4.706666666666667e-05, "loss": 3.1673828125, "step": 220 }, { "epoch": 0.176, "model.model.language_model.layers.26.mlp.down_proj": 0.486328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49951171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.6025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.886474609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.193603515625, "step": 220 }, { "epoch": 0.1768, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49853515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.60205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.190185546875, "step": 221 }, { "epoch": 0.1776, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.416015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.498046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.72021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.601318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.189697265625, "step": 222 }, { "epoch": 0.1784, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.345458984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.51171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.400390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.497802734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.60009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.51953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.18896484375, "step": 223 }, { "epoch": 0.1792, "model.model.language_model.layers.26.mlp.down_proj": 0.53125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3447265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823974609375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4970703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.598876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885986328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1884765625, "step": 224 }, { "epoch": 0.18, "model.model.language_model.layers.26.mlp.down_proj": 0.4765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.343994140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.44140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.496337890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.59765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885986328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.186767578125, "step": 225 }, { "epoch": 0.1808, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.34326171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.495361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.597412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.185791015625, "step": 226 }, { "epoch": 0.1816, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.34228515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.59619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.40625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.184814453125, "step": 227 }, { "epoch": 0.1824, "model.model.language_model.layers.26.mlp.down_proj": 0.55078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.34228515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.494140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.404296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.493896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.595458984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.184326171875, "step": 228 }, { "epoch": 0.1832, "model.model.language_model.layers.26.mlp.down_proj": 0.53125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.340576171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.51171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7197265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.59521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.18310546875, "step": 229 }, { "epoch": 0.184, "grad_norm": 0.3010888397693634, "learning_rate": 4.6933333333333333e-05, "loss": 3.1708984375, "step": 230 }, { "epoch": 0.184, "model.model.language_model.layers.26.mlp.down_proj": 0.54296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.33935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.46484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.492919921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.593505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.181884765625, "step": 230 }, { "epoch": 0.1848, "model.model.language_model.layers.26.mlp.down_proj": 0.5546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.337646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.494140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.408203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49267578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.56640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.59326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.18115234375, "step": 231 }, { "epoch": 0.1856, "model.model.language_model.layers.26.mlp.down_proj": 0.474609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.336181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.823486328125, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.491455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.718994140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.592041015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.179931640625, "step": 232 }, { "epoch": 0.1864, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3349609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.474609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.490966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.718994140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.46875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.591552734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1787109375, "step": 233 }, { "epoch": 0.1872, "model.model.language_model.layers.26.mlp.down_proj": 0.56640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.334716796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.4140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.490966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.591064453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.177978515625, "step": 234 }, { "epoch": 0.188, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.334228515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.44140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.718505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.17724609375, "step": 235 }, { "epoch": 0.1888, "model.model.language_model.layers.26.mlp.down_proj": 0.4453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.33349609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.462890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.49072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.885009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.17724609375, "step": 236 }, { "epoch": 0.1896, "model.model.language_model.layers.26.mlp.down_proj": 0.47265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.332275390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.490478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.45703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.589111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.884521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.17626953125, "step": 237 }, { "epoch": 0.1904, "model.model.language_model.layers.26.mlp.down_proj": 0.48828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.331787109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.482421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.490478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.718017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.482421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.588623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.175537109375, "step": 238 }, { "epoch": 0.1912, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3310546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.470703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48974609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.718017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.5, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.587890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.174560546875, "step": 239 }, { "epoch": 0.192, "grad_norm": 0.28878313302993774, "learning_rate": 4.6800000000000006e-05, "loss": 3.113671875, "step": 240 }, { "epoch": 0.192, "model.model.language_model.layers.26.mlp.down_proj": 0.5390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.329345703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822509765625, "model.model.language_model.layers.26.mlp.up_proj": 0.349609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48876953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7177734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.587158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.17236328125, "step": 240 }, { "epoch": 0.1928, "model.model.language_model.layers.26.mlp.down_proj": 0.451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.478515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822509765625, "model.model.language_model.layers.26.mlp.up_proj": 0.359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.488525390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7177734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.586669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.40625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.172119140625, "step": 241 }, { "epoch": 0.1936, "model.model.language_model.layers.26.mlp.down_proj": 0.494140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.32763671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.45703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822509765625, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.488525390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.717529296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.171875, "step": 242 }, { "epoch": 0.1944, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3271484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.443359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.34765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.586181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.171630859375, "step": 243 }, { "epoch": 0.1952, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.474609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.384765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.171142578125, "step": 244 }, { "epoch": 0.196, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3251953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.46875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48779296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.717041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.170166015625, "step": 245 }, { "epoch": 0.1968, "model.model.language_model.layers.26.mlp.down_proj": 0.47265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.324951171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.427734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48779296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.584716796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.16943359375, "step": 246 }, { "epoch": 0.1976, "model.model.language_model.layers.26.mlp.down_proj": 0.52734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.32470703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.494140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.40234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48779296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.55078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58447265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1689453125, "step": 247 }, { "epoch": 0.1984, "model.model.language_model.layers.26.mlp.down_proj": 0.455078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.32373046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.487548828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.583740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.167724609375, "step": 248 }, { "epoch": 0.1992, "model.model.language_model.layers.26.mlp.down_proj": 0.486328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4873046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58349609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.167236328125, "step": 249 }, { "epoch": 0.2, "grad_norm": 0.32630211114883423, "learning_rate": 4.666666666666667e-05, "loss": 3.0341796875, "step": 250 }, { "epoch": 0.2, "model.model.language_model.layers.26.mlp.down_proj": 0.60546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.52734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.455078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4873046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.52734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.166748046875, "step": 250 }, { "epoch": 0.2008, "model.model.language_model.layers.26.mlp.down_proj": 0.451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.320068359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.439453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.34765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.487060546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.582763671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.166259765625, "step": 251 }, { "epoch": 0.2016, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.318603515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48681640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.58251953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.166015625, "step": 252 }, { "epoch": 0.2024, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.486083984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5810546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.164794921875, "step": 253 }, { "epoch": 0.2032, "model.model.language_model.layers.26.mlp.down_proj": 0.486328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.317138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.49609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.485595703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.580322265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.164306640625, "step": 254 }, { "epoch": 0.204, "model.model.language_model.layers.26.mlp.down_proj": 0.49609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.315673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.478515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4853515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.579345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.163818359375, "step": 255 }, { "epoch": 0.2048, "model.model.language_model.layers.26.mlp.down_proj": 0.486328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48486328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.716064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.577880859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.883544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.163818359375, "step": 256 }, { "epoch": 0.2056, "model.model.language_model.layers.26.mlp.down_proj": 0.47265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.313720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.447265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.577392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.16259765625, "step": 257 }, { "epoch": 0.2064, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.31298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.42578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.484130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.57666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.162109375, "step": 258 }, { "epoch": 0.2072, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.311767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4833984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.5, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.576171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.160888671875, "step": 259 }, { "epoch": 0.208, "grad_norm": 0.24819865822792053, "learning_rate": 4.653333333333334e-05, "loss": 2.978125, "step": 260 }, { "epoch": 0.208, "model.model.language_model.layers.26.mlp.down_proj": 0.51171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.311279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.482666015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.575927734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.160400390625, "step": 260 }, { "epoch": 0.2088, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.380859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.34375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.159912109375, "step": 261 }, { "epoch": 0.2096, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.310302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.439453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82177734375, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.482177734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.34375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.574951171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.15966796875, "step": 262 }, { "epoch": 0.2104, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.30908203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.419921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.821533203125, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.573974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.159423828125, "step": 263 }, { "epoch": 0.2112, "model.model.language_model.layers.26.mlp.down_proj": 0.58984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.388671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.573974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.157958984375, "step": 264 }, { "epoch": 0.212, "model.model.language_model.layers.26.mlp.down_proj": 0.490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3076171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.427734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.48046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.572509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.157470703125, "step": 265 }, { "epoch": 0.2128, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.439453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.480224609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.571533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.156982421875, "step": 266 }, { "epoch": 0.2136, "model.model.language_model.layers.26.mlp.down_proj": 0.44921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.305908203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.821044921875, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47998046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.571533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.156982421875, "step": 267 }, { "epoch": 0.2144, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3056640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.48046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.821044921875, "model.model.language_model.layers.26.mlp.up_proj": 0.37890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.479248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.15673828125, "step": 268 }, { "epoch": 0.2152, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.30517578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.416015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.82080078125, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.479248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.57080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.156494140625, "step": 269 }, { "epoch": 0.216, "grad_norm": 0.27884969115257263, "learning_rate": 4.64e-05, "loss": 2.9548828125, "step": 270 }, { "epoch": 0.216, "model.model.language_model.layers.26.mlp.down_proj": 0.447265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.820556640625, "model.model.language_model.layers.26.mlp.up_proj": 0.337890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47900390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.715087890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.156005859375, "step": 270 }, { "epoch": 0.2168, "model.model.language_model.layers.26.mlp.down_proj": 0.48828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.303466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.455078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.820556640625, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.478759765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.569580078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.155517578125, "step": 271 }, { "epoch": 0.2176, "model.model.language_model.layers.26.mlp.down_proj": 0.546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.30322265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.50390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8203125, "model.model.language_model.layers.26.mlp.up_proj": 0.400390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47802734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.714599609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.568603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.155029296875, "step": 272 }, { "epoch": 0.2184, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.302978515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8203125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47802734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.714599609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.568115234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.154052734375, "step": 273 }, { "epoch": 0.2192, "model.model.language_model.layers.26.mlp.down_proj": 0.47265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.302978515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.46875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8203125, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47802734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.714599609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.567626953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.466796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.154052734375, "step": 274 }, { "epoch": 0.22, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8203125, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.477783203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71435546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.567626953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.153076171875, "step": 275 }, { "epoch": 0.2208, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.302001953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.474609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8203125, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.477783203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.714111328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.567138671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.153076171875, "step": 276 }, { "epoch": 0.2216, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.301025390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47705078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.56591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.15234375, "step": 277 }, { "epoch": 0.2224, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.476806640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882568359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1513671875, "step": 278 }, { "epoch": 0.2232, "model.model.language_model.layers.26.mlp.down_proj": 0.482421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.299560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.466796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.56396484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.482421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882568359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.151123046875, "step": 279 }, { "epoch": 0.224, "grad_norm": 0.27415135502815247, "learning_rate": 4.626666666666667e-05, "loss": 2.9005859375, "step": 280 }, { "epoch": 0.224, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29931640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.46484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.3828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.476318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.51171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882568359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.151123046875, "step": 280 }, { "epoch": 0.2248, "model.model.language_model.layers.26.mlp.down_proj": 0.5234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.298583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.48828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.476318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.53125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.56298828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882568359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.149658203125, "step": 281 }, { "epoch": 0.2256, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2978515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.416015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.819091796875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.476318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.56201171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1494140625, "step": 282 }, { "epoch": 0.2264, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.297607421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.819091796875, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.476318359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713623046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.49609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.561767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1494140625, "step": 283 }, { "epoch": 0.2272, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71337890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5615234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.14892578125, "step": 284 }, { "epoch": 0.228, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29638671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.380859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71337890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.560302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.148681640625, "step": 285 }, { "epoch": 0.2288, "model.model.language_model.layers.26.mlp.down_proj": 0.45703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2958984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.458984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.475830078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71337890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.559814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.148193359375, "step": 286 }, { "epoch": 0.2296, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.295654296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.419921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.475830078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71337890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.559326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.148193359375, "step": 287 }, { "epoch": 0.2304, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29541015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.337890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.475830078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71337890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.55908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.147705078125, "step": 288 }, { "epoch": 0.2312, "model.model.language_model.layers.26.mlp.down_proj": 0.455078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.295166015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.455078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4755859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.713134765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.558837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.882080078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.147216796875, "step": 289 }, { "epoch": 0.232, "grad_norm": 0.3102974593639374, "learning_rate": 4.6133333333333334e-05, "loss": 2.8765625, "step": 290 }, { "epoch": 0.232, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.294677734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47509765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.712646484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.55810546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8818359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.14599609375, "step": 290 }, { "epoch": 0.2328, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.294189453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.357421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.474853515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.712646484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.557861328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8818359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.145751953125, "step": 291 }, { "epoch": 0.2336, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29345703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.41796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.818603515625, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.474365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.712646484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.556396484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.881591796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.145751953125, "step": 292 }, { "epoch": 0.2344, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.408203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.818115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47314453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.712158203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.55419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.498046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.881591796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.143798828125, "step": 293 }, { "epoch": 0.2352, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.29150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.447265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.818115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.472412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.712158203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5537109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.881591796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1435546875, "step": 294 }, { "epoch": 0.236, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.472412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5537109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.88134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.142822265625, "step": 295 }, { "epoch": 0.2368, "model.model.language_model.layers.26.mlp.down_proj": 0.48828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.290771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.455078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47216796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.711669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.55224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.881103515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.141845703125, "step": 296 }, { "epoch": 0.2376, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.290771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.44921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.3671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47216796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.711669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.48828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.551513671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.880859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1416015625, "step": 297 }, { "epoch": 0.2384, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28857421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4716796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.711669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.550048828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8798828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.139404296875, "step": 298 }, { "epoch": 0.2392, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2880859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4716796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.711669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.549560546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.879638671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.138916015625, "step": 299 }, { "epoch": 0.24, "grad_norm": 0.30015042424201965, "learning_rate": 4.600000000000001e-05, "loss": 2.810546875, "step": 300 }, { "epoch": 0.24, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28759765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.44921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.34765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.549072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87939453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.138671875, "step": 300 }, { "epoch": 0.2408, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28662109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.47021484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87939453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.138427734375, "step": 301 }, { "epoch": 0.2416, "model.model.language_model.layers.26.mlp.down_proj": 0.458984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.286376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.4609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.817626953125, "model.model.language_model.layers.26.mlp.up_proj": 0.369140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.469970703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.548583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.879150390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.137939453125, "step": 302 }, { "epoch": 0.2424, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.286376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.486328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.817626953125, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.469970703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.48046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.54833984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.879150390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.137451171875, "step": 303 }, { "epoch": 0.2432, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.817626953125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4697265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.547607421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.879150390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.13720703125, "step": 304 }, { "epoch": 0.244, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.817138671875, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.469482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.546630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.879150390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.136474609375, "step": 305 }, { "epoch": 0.2448, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.284423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.419921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.817138671875, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46923828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.546630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.13623046875, "step": 306 }, { "epoch": 0.2456, "model.model.language_model.layers.26.mlp.down_proj": 0.408203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81689453125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.546142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.13623046875, "step": 307 }, { "epoch": 0.2464, "model.model.language_model.layers.26.mlp.down_proj": 0.470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.283935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81689453125, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5458984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.135986328125, "step": 308 }, { "epoch": 0.2472, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28369140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.40234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81689453125, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.71142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.545654296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1357421875, "step": 309 }, { "epoch": 0.248, "grad_norm": 0.2859407961368561, "learning_rate": 4.5866666666666666e-05, "loss": 2.781640625, "step": 310 }, { "epoch": 0.248, "model.model.language_model.layers.26.mlp.down_proj": 0.47265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28369140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81689453125, "model.model.language_model.layers.26.mlp.up_proj": 0.345703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.711181640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.474609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.135009765625, "step": 310 }, { "epoch": 0.2488, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28369140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81689453125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.54443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.134521484375, "step": 311 }, { "epoch": 0.2496, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.283447265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.447265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.816650390625, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.468017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5439453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1337890625, "step": 312 }, { "epoch": 0.2504, "model.model.language_model.layers.26.mlp.down_proj": 0.5, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.283447265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.46875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.816650390625, "model.model.language_model.layers.26.mlp.up_proj": 0.37109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.482421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.543701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1328125, "step": 313 }, { "epoch": 0.2512, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.816650390625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.542724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.132080078125, "step": 314 }, { "epoch": 0.252, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.281494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46728515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.542236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1318359375, "step": 315 }, { "epoch": 0.2528, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj": 0.490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.408203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46728515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.53125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.542236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.131591796875, "step": 316 }, { "epoch": 0.2536, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.466552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.541748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.131103515625, "step": 317 }, { "epoch": 0.2544, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.466552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.541259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.130859375, "step": 318 }, { "epoch": 0.2552, "model.model.language_model.layers.26.mlp.down_proj": 0.46875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.45703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.54052734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.129638671875, "step": 319 }, { "epoch": 0.256, "grad_norm": 0.2341630905866623, "learning_rate": 4.573333333333333e-05, "loss": 2.79140625, "step": 320 }, { "epoch": 0.256, "model.model.language_model.layers.26.mlp.down_proj": 0.484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.279052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.443359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81640625, "model.model.language_model.layers.26.mlp.up_proj": 0.36328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.54052734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12939453125, "step": 320 }, { "epoch": 0.2568, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.278564453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.816162109375, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.466064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12939453125, "step": 321 }, { "epoch": 0.2576, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2783203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.427734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.815673828125, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.466064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.539794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.129150390625, "step": 322 }, { "epoch": 0.2584, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.47265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.815185546875, "model.model.language_model.layers.26.mlp.up_proj": 0.38671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4658203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.710205078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12890625, "step": 323 }, { "epoch": 0.2592, "model.model.language_model.layers.26.mlp.down_proj": 0.49609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.277099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.466796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.815185546875, "model.model.language_model.layers.26.mlp.up_proj": 0.3828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46533203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7099609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.53857421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12890625, "step": 324 }, { "epoch": 0.26, "model.model.language_model.layers.26.mlp.down_proj": 0.470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27685546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46533203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7099609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.53857421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.128662109375, "step": 325 }, { "epoch": 0.2608, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.276611328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.4140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.464599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7099609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.538330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1279296875, "step": 326 }, { "epoch": 0.2616, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.276123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.427734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7099609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.127685546875, "step": 327 }, { "epoch": 0.2624, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814697265625, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.709716796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.536865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12646484375, "step": 328 }, { "epoch": 0.2632, "model.model.language_model.layers.26.mlp.down_proj": 0.48828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.275146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.416015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.709716796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.535400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.124267578125, "step": 329 }, { "epoch": 0.264, "grad_norm": 0.24989570677280426, "learning_rate": 4.5600000000000004e-05, "loss": 2.755859375, "step": 330 }, { "epoch": 0.264, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.709716796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.53515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878173828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.123779296875, "step": 330 }, { "epoch": 0.2648, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.274169921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70947265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.534912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.878173828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.123291015625, "step": 331 }, { "epoch": 0.2656, "model.model.language_model.layers.26.mlp.down_proj": 0.455078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27392578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.337890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70947265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.534423828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8779296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.122802734375, "step": 332 }, { "epoch": 0.2664, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.273193359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.463623046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.709228515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.53271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8779296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.121826171875, "step": 333 }, { "epoch": 0.2672, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.463623046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.709228515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5322265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8779296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12158203125, "step": 334 }, { "epoch": 0.268, "model.model.language_model.layers.26.mlp.down_proj": 0.455078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.53173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8779296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12158203125, "step": 335 }, { "epoch": 0.2688, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.272216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708740234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.530517578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87744140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12158203125, "step": 336 }, { "epoch": 0.2696, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.3359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708740234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.45703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.530517578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.877197265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.121337890625, "step": 337 }, { "epoch": 0.2704, "model.model.language_model.layers.26.mlp.down_proj": 0.470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708740234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.530029296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.877197265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.12109375, "step": 338 }, { "epoch": 0.2712, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.27099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4619140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708740234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.528076171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.120849609375, "step": 339 }, { "epoch": 0.272, "grad_norm": 0.27415409684181213, "learning_rate": 4.546666666666667e-05, "loss": 2.720703125, "step": 340 }, { "epoch": 0.272, "model.model.language_model.layers.26.mlp.down_proj": 0.5234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.269287109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.34765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.461669921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708251953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.526611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.47265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.876708984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.119140625, "step": 340 }, { "epoch": 0.2728, "model.model.language_model.layers.26.mlp.down_proj": 0.48828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26806640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.461669921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708251953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.525634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.118896484375, "step": 341 }, { "epoch": 0.2736, "model.model.language_model.layers.26.mlp.down_proj": 0.408203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.461669921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.708251953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.525146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.118896484375, "step": 342 }, { "epoch": 0.2744, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.267333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7080078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.524169921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.876220703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.117919921875, "step": 343 }, { "epoch": 0.2752, "model.model.language_model.layers.26.mlp.down_proj": 0.447265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.707763671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.523681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.117919921875, "step": 344 }, { "epoch": 0.276, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.44921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70751953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.523681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11767578125, "step": 345 }, { "epoch": 0.2768, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.266845703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70751953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.523193359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11767578125, "step": 346 }, { "epoch": 0.2776, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2666015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.408203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.461181640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70751953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.52294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.117431640625, "step": 347 }, { "epoch": 0.2784, "model.model.language_model.layers.26.mlp.down_proj": 0.478515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2666015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70751953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.522216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.117431640625, "step": 348 }, { "epoch": 0.2792, "model.model.language_model.layers.26.mlp.down_proj": 0.453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2666015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.42578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.3359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70751953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.521728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.117431640625, "step": 349 }, { "epoch": 0.28, "grad_norm": 0.24260057508945465, "learning_rate": 4.5333333333333335e-05, "loss": 2.703125, "step": 350 }, { "epoch": 0.28, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26611328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.707275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1171875, "step": 350 }, { "epoch": 0.2808, "model.model.language_model.layers.26.mlp.down_proj": 0.48046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.466796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.707275390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.520751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.116943359375, "step": 351 }, { "epoch": 0.2816, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.52001953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.116455078125, "step": 352 }, { "epoch": 0.2824, "model.model.language_model.layers.26.mlp.down_proj": 0.45703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.4453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.35546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875732421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1162109375, "step": 353 }, { "epoch": 0.2832, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.518798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.115966796875, "step": 354 }, { "epoch": 0.284, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.263916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.814208984375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.518310546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.115234375, "step": 355 }, { "epoch": 0.2848, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81396484375, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.115234375, "step": 356 }, { "epoch": 0.2856, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.263427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.458984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81396484375, "model.model.language_model.layers.26.mlp.up_proj": 0.373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.494140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.40625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.115234375, "step": 357 }, { "epoch": 0.2864, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81396484375, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.517333984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1142578125, "step": 358 }, { "epoch": 0.2872, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.261962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.46044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.113525390625, "step": 359 }, { "epoch": 0.288, "grad_norm": 0.21247848868370056, "learning_rate": 4.52e-05, "loss": 2.6861328125, "step": 360 }, { "epoch": 0.288, "model.model.language_model.layers.26.mlp.down_proj": 0.51171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.478515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.39453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.435546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.113525390625, "step": 360 }, { "epoch": 0.2888, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2607421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.514892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.113037109375, "step": 361 }, { "epoch": 0.2896, "model.model.language_model.layers.26.mlp.down_proj": 0.462890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2607421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.44140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.462890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.113037109375, "step": 362 }, { "epoch": 0.2904, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.260498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11279296875, "step": 363 }, { "epoch": 0.2912, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.26025390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11279296875, "step": 364 }, { "epoch": 0.292, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.260009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813720703125, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.459716796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.112548828125, "step": 365 }, { "epoch": 0.2928, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45947265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.513427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875244140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.112060546875, "step": 366 }, { "epoch": 0.2936, "model.model.language_model.layers.26.mlp.down_proj": 0.423828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.258544921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.423828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45947265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875244140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.111572265625, "step": 367 }, { "epoch": 0.2944, "model.model.language_model.layers.26.mlp.down_proj": 0.447265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.258056640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.447265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45947265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875244140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.111572265625, "step": 368 }, { "epoch": 0.2952, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.258056640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.435546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.512939453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875, "model.model.language_model.layers.26.self_attn.v_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.111083984375, "step": 369 }, { "epoch": 0.296, "grad_norm": 0.2618201971054077, "learning_rate": 4.5066666666666667e-05, "loss": 2.6353515625, "step": 370 }, { "epoch": 0.296, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.512939453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875, "model.model.language_model.layers.26.self_attn.v_proj": 0.359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11083984375, "step": 370 }, { "epoch": 0.2968, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.458740234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.11083984375, "step": 371 }, { "epoch": 0.2976, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.257568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874755859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1103515625, "step": 372 }, { "epoch": 0.2984, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874755859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1103515625, "step": 373 }, { "epoch": 0.2992, "model.model.language_model.layers.26.mlp.down_proj": 0.4296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.510498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87451171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.110107421875, "step": 374 }, { "epoch": 0.3, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.25634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.458251953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874267578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10986328125, "step": 375 }, { "epoch": 0.3008, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.462890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8134765625, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874267578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10986328125, "step": 376 }, { "epoch": 0.3016, "model.model.language_model.layers.26.mlp.down_proj": 0.458984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.41796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.457763671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.51025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874267578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10986328125, "step": 377 }, { "epoch": 0.3024, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45751953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.510009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874267578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.109619140625, "step": 378 }, { "epoch": 0.3032, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.255615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45751953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.874267578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.108642578125, "step": 379 }, { "epoch": 0.304, "grad_norm": 0.2572396993637085, "learning_rate": 4.493333333333333e-05, "loss": 2.6392578125, "step": 380 }, { "epoch": 0.304, "model.model.language_model.layers.26.mlp.down_proj": 0.447265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45751953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.509521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.108154296875, "step": 380 }, { "epoch": 0.3048, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.25439453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45751953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.509033203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.108154296875, "step": 381 }, { "epoch": 0.3056, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.254150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.457275390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.508544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.108154296875, "step": 382 }, { "epoch": 0.3064, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.254150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.813232421875, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.457275390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.108154296875, "step": 383 }, { "epoch": 0.3072, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.254150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.457275390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.508056640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10791015625, "step": 384 }, { "epoch": 0.308, "model.model.language_model.layers.26.mlp.down_proj": 0.423828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.254150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.706298828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.507568359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.107666015625, "step": 385 }, { "epoch": 0.3088, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.253173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456787109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5068359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873779296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.107177734375, "step": 386 }, { "epoch": 0.3096, "model.model.language_model.layers.26.mlp.down_proj": 0.4296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.252685546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456787109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87353515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.107177734375, "step": 387 }, { "epoch": 0.3104, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.252685546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.506103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87353515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10693359375, "step": 388 }, { "epoch": 0.3112, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.25244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.40234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50537109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87353515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.106689453125, "step": 389 }, { "epoch": 0.312, "grad_norm": 0.23896296322345734, "learning_rate": 4.4800000000000005e-05, "loss": 2.6150390625, "step": 390 }, { "epoch": 0.312, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.252197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.505126953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87353515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.106689453125, "step": 390 }, { "epoch": 0.3128, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.251708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50439453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87353515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.106201171875, "step": 391 }, { "epoch": 0.3136, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.250732421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50439453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873291015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.105712890625, "step": 392 }, { "epoch": 0.3144, "model.model.language_model.layers.26.mlp.down_proj": 0.44921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.250732421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4560546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.439453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.504150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873291015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10546875, "step": 393 }, { "epoch": 0.3152, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.250244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.41796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4560546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.105224609375, "step": 394 }, { "epoch": 0.316, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24951171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8125, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455810546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.503662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10498046875, "step": 395 }, { "epoch": 0.3168, "model.model.language_model.layers.26.mlp.down_proj": 0.4765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.249267578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.443359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.812255859375, "model.model.language_model.layers.26.mlp.up_proj": 0.357421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455810546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.503173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10498046875, "step": 396 }, { "epoch": 0.3176, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81201171875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455810546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.502685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.872802734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10498046875, "step": 397 }, { "epoch": 0.3184, "model.model.language_model.layers.26.mlp.down_proj": 0.451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.248779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81201171875, "model.model.language_model.layers.26.mlp.up_proj": 0.369140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.3125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.872802734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10498046875, "step": 398 }, { "epoch": 0.3192, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24853515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.42578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81201171875, "model.model.language_model.layers.26.mlp.up_proj": 0.341796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.872802734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.104736328125, "step": 399 }, { "epoch": 0.32, "grad_norm": 0.2071940153837204, "learning_rate": 4.466666666666667e-05, "loss": 2.5970703125, "step": 400 }, { "epoch": 0.32, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.247802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81201171875, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.872802734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10400390625, "step": 400 }, { "epoch": 0.3208, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.501708984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10400390625, "step": 401 }, { "epoch": 0.3216, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.247314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.419921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10400390625, "step": 402 }, { "epoch": 0.3224, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24658203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.357421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.4765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10400390625, "step": 403 }, { "epoch": 0.3232, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.246337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.50146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10400390625, "step": 404 }, { "epoch": 0.324, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.246337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705322265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.5009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.103271484375, "step": 405 }, { "epoch": 0.3248, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.245849609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705322265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.500244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.103271484375, "step": 406 }, { "epoch": 0.3256, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705322265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.499267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.103271484375, "step": 407 }, { "epoch": 0.3264, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.103271484375, "step": 408 }, { "epoch": 0.3272, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.404296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.497802734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.102783203125, "step": 409 }, { "epoch": 0.328, "grad_norm": 0.3187468349933624, "learning_rate": 4.4533333333333336e-05, "loss": 2.57265625, "step": 410 }, { "epoch": 0.328, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.244873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1025390625, "step": 410 }, { "epoch": 0.3288, "model.model.language_model.layers.26.mlp.down_proj": 0.4296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24462890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.408203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811767578125, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45458984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4970703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1025390625, "step": 411 }, { "epoch": 0.3296, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.243896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.454345703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.101806640625, "step": 412 }, { "epoch": 0.3304, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.243896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.454345703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.435546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87255859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.101806640625, "step": 413 }, { "epoch": 0.3312, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.243408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.705078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.872314453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.101806640625, "step": 414 }, { "epoch": 0.332, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.243408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8115234375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10107421875, "step": 415 }, { "epoch": 0.3328, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24267578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811279296875, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.100830078125, "step": 416 }, { "epoch": 0.3336, "model.model.language_model.layers.26.mlp.down_proj": 0.48046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811279296875, "model.model.language_model.layers.26.mlp.up_proj": 0.349609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453857421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.50390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.495849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.1005859375, "step": 417 }, { "epoch": 0.3344, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.241943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.404296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811279296875, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453857421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49560546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.100341796875, "step": 418 }, { "epoch": 0.3352, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.241455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811279296875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4951171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10009765625, "step": 419 }, { "epoch": 0.336, "grad_norm": 0.22022688388824463, "learning_rate": 4.44e-05, "loss": 2.5416015625, "step": 420 }, { "epoch": 0.336, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.241455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.404296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.811279296875, "model.model.language_model.layers.26.mlp.up_proj": 0.333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8720703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.10009765625, "step": 420 }, { "epoch": 0.3368, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.241455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81103515625, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871826171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.099853515625, "step": 421 }, { "epoch": 0.3376, "model.model.language_model.layers.26.mlp.down_proj": 0.49609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.241455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81103515625, "model.model.language_model.layers.26.mlp.up_proj": 0.384765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704833984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.493896484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871826171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.099853515625, "step": 422 }, { "epoch": 0.3384, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.439453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81103515625, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45361328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70458984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4931640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871826171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.099609375, "step": 423 }, { "epoch": 0.3392, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.24072265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.810302734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70458984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.098876953125, "step": 424 }, { "epoch": 0.34, "model.model.language_model.layers.26.mlp.down_proj": 0.5078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.44921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.491943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.098876953125, "step": 425 }, { "epoch": 0.3408, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49169921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.098876953125, "step": 426 }, { "epoch": 0.3416, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871337890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.097900390625, "step": 427 }, { "epoch": 0.3424, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.490966796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871337890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09765625, "step": 428 }, { "epoch": 0.3432, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.46484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.49072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871337890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.097412109375, "step": 429 }, { "epoch": 0.344, "grad_norm": 0.1952962428331375, "learning_rate": 4.426666666666667e-05, "loss": 2.532421875, "step": 430 }, { "epoch": 0.344, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.239990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.42578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.489990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871337890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.097412109375, "step": 430 }, { "epoch": 0.3448, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23974609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.871337890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09716796875, "step": 431 }, { "epoch": 0.3456, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.239501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.489501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.096923828125, "step": 432 }, { "epoch": 0.3464, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.81005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.096923828125, "step": 433 }, { "epoch": 0.3472, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23876953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.809814453125, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.704345703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.096923828125, "step": 434 }, { "epoch": 0.348, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.236572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.809326171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.488037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0966796875, "step": 435 }, { "epoch": 0.3488, "model.model.language_model.layers.26.mlp.down_proj": 0.48046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.4375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.809326171875, "model.model.language_model.layers.26.mlp.up_proj": 0.361328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.488037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.41015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0966796875, "step": 436 }, { "epoch": 0.3496, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.487548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.095947265625, "step": 437 }, { "epoch": 0.3504, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.095947265625, "step": 438 }, { "epoch": 0.3512, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.095703125, "step": 439 }, { "epoch": 0.352, "grad_norm": 0.2249034345149994, "learning_rate": 4.413333333333334e-05, "loss": 2.4734375, "step": 440 }, { "epoch": 0.352, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.87060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09521484375, "step": 440 }, { "epoch": 0.3528, "model.model.language_model.layers.26.mlp.down_proj": 0.455078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.41796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.34765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.486328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.487060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.870361328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09521484375, "step": 441 }, { "epoch": 0.3536, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.234130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.870361328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09521484375, "step": 442 }, { "epoch": 0.3544, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.870361328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094970703125, "step": 443 }, { "epoch": 0.3552, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.452392578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.486083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.870361328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094970703125, "step": 444 }, { "epoch": 0.356, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808837890625, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.485107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.870361328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0947265625, "step": 445 }, { "epoch": 0.3568, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808837890625, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0947265625, "step": 446 }, { "epoch": 0.3576, "model.model.language_model.layers.26.mlp.down_proj": 0.4453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.408203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808837890625, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094482421875, "step": 447 }, { "epoch": 0.3584, "model.model.language_model.layers.26.mlp.down_proj": 0.4609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.427734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808837890625, "model.model.language_model.layers.26.mlp.up_proj": 0.349609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094482421875, "step": 448 }, { "epoch": 0.3592, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.232666015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80859375, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094482421875, "step": 449 }, { "epoch": 0.36, "grad_norm": 0.21266672015190125, "learning_rate": 4.4000000000000006e-05, "loss": 2.49453125, "step": 450 }, { "epoch": 0.36, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80859375, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094482421875, "step": 450 }, { "epoch": 0.3608, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.232177734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80859375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.484130859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8701171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.094482421875, "step": 451 }, { "epoch": 0.3616, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23193359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.093994140625, "step": 452 }, { "epoch": 0.3624, "model.model.language_model.layers.26.mlp.down_proj": 0.51171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.443359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.376953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09375, "step": 453 }, { "epoch": 0.3632, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.231201171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.443359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09375, "step": 454 }, { "epoch": 0.364, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.231201171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09375, "step": 455 }, { "epoch": 0.3648, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.231201171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.808349609375, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09375, "step": 456 }, { "epoch": 0.3656, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23095703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.093505859375, "step": 457 }, { "epoch": 0.3664, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.23095703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80810546875, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09326171875, "step": 458 }, { "epoch": 0.3672, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.230712890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869873046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.093017578125, "step": 459 }, { "epoch": 0.368, "grad_norm": 0.1950872242450714, "learning_rate": 4.3866666666666665e-05, "loss": 2.470703125, "step": 460 }, { "epoch": 0.368, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.229736328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.482177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86962890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.093017578125, "step": 460 }, { "epoch": 0.3688, "model.model.language_model.layers.26.mlp.down_proj": 0.435546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.229736328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.349609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.481689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86962890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.093017578125, "step": 461 }, { "epoch": 0.3696, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.229736328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.481689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86962890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0927734375, "step": 462 }, { "epoch": 0.3704, "model.model.language_model.layers.26.mlp.down_proj": 0.4453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.349609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869384765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0927734375, "step": 463 }, { "epoch": 0.3712, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.228759765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.48095703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.092041015625, "step": 464 }, { "epoch": 0.372, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.228271484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.451171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703857421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.479736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.869140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.091552734375, "step": 465 }, { "epoch": 0.3728, "model.model.language_model.layers.26.mlp.down_proj": 0.5234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.227294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.462890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.3984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.450927734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70361328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.427734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.479248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.090576171875, "step": 466 }, { "epoch": 0.3736, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.227294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.450927734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.479248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.090576171875, "step": 467 }, { "epoch": 0.3744, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.227294921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45068359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.479248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09033203125, "step": 468 }, { "epoch": 0.3752, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.45068359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09033203125, "step": 469 }, { "epoch": 0.376, "grad_norm": 0.2537761926651001, "learning_rate": 4.373333333333334e-05, "loss": 2.4478515625, "step": 470 }, { "epoch": 0.376, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.226806640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.450439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.478759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.09033203125, "step": 470 }, { "epoch": 0.3768, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.226318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.450439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.478759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.090087890625, "step": 471 }, { "epoch": 0.3776, "model.model.language_model.layers.26.mlp.down_proj": 0.470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.226318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.423828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.450439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.4375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.478759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.089599609375, "step": 472 }, { "epoch": 0.3784, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.225830078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.089599609375, "step": 473 }, { "epoch": 0.3792, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.225830078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.868896484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.089599609375, "step": 474 }, { "epoch": 0.38, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.225341796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47705078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.089111328125, "step": 475 }, { "epoch": 0.3808, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22509765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0888671875, "step": 476 }, { "epoch": 0.3816, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.224853515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.449951171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0888671875, "step": 477 }, { "epoch": 0.3824, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.224853515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.449951171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47607421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0888671875, "step": 478 }, { "epoch": 0.3832, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.224609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.449462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47607421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0888671875, "step": 479 }, { "epoch": 0.384, "grad_norm": 0.1995895504951477, "learning_rate": 4.36e-05, "loss": 2.415234375, "step": 480 }, { "epoch": 0.384, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.224609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08837890625, "step": 480 }, { "epoch": 0.3848, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.40234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8681640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08837890625, "step": 481 }, { "epoch": 0.3856, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.448974609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867919921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.087890625, "step": 482 }, { "epoch": 0.3864, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.448486328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867919921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.087646484375, "step": 483 }, { "epoch": 0.3872, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.223876953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.448486328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.474853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867919921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08740234375, "step": 484 }, { "epoch": 0.388, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.223876953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.474853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867919921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08740234375, "step": 485 }, { "epoch": 0.3888, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2236328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08740234375, "step": 486 }, { "epoch": 0.3896, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.223388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447998046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08740234375, "step": 487 }, { "epoch": 0.3904, "model.model.language_model.layers.26.mlp.down_proj": 0.435546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.223388671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.41015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447998046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08740234375, "step": 488 }, { "epoch": 0.3912, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.222900390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.807373046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447998046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0869140625, "step": 489 }, { "epoch": 0.392, "grad_norm": 0.2101357877254486, "learning_rate": 4.346666666666667e-05, "loss": 2.3884765625, "step": 490 }, { "epoch": 0.392, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80712890625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44775390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0869140625, "step": 490 }, { "epoch": 0.3928, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80712890625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44775390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.473388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0869140625, "step": 491 }, { "epoch": 0.3936, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.222412109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80712890625, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44775390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0869140625, "step": 492 }, { "epoch": 0.3944, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80712890625, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44775390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0869140625, "step": 493 }, { "epoch": 0.3952, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.221923828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447509765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86767578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.086669921875, "step": 494 }, { "epoch": 0.396, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447509765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.472412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08642578125, "step": 495 }, { "epoch": 0.3968, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447509765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08642578125, "step": 496 }, { "epoch": 0.3976, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.221435546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806884765625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08642578125, "step": 497 }, { "epoch": 0.3984, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806640625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.086181640625, "step": 498 }, { "epoch": 0.3992, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22021484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.806396484375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447021484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47119140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.086181640625, "step": 499 }, { "epoch": 0.4, "grad_norm": 0.22104111313819885, "learning_rate": 4.3333333333333334e-05, "loss": 2.36328125, "step": 500 }, { "epoch": 0.4, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.22021484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.447021484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.470947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0859375, "step": 500 }, { "epoch": 0.4008, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.21923828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0859375, "step": 501 }, { "epoch": 0.4016, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.218505859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.47021484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0859375, "step": 502 }, { "epoch": 0.4024, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.218017578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085693359375, "step": 503 }, { "epoch": 0.4032, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.217529296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703369140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.469482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085693359375, "step": 504 }, { "epoch": 0.404, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.217529296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80615234375, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085693359375, "step": 505 }, { "epoch": 0.4048, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.216064453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.805908203125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.446044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 506 }, { "epoch": 0.4056, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.215576171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8056640625, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.446044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 507 }, { "epoch": 0.4064, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.215576171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.805419921875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.446044921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 508 }, { "epoch": 0.4072, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.215576171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.805419921875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 509 }, { "epoch": 0.408, "grad_norm": 0.21670912206172943, "learning_rate": 4.32e-05, "loss": 2.3798828125, "step": 510 }, { "epoch": 0.408, "model.model.language_model.layers.26.mlp.down_proj": 0.44140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.21533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.42578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.805419921875, "model.model.language_model.layers.26.mlp.up_proj": 0.365234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.447265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 510 }, { "epoch": 0.4088, "model.model.language_model.layers.26.mlp.down_proj": 0.408203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.21533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.805419921875, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.867431640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 511 }, { "epoch": 0.4096, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.215087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.085205078125, "step": 512 }, { "epoch": 0.4104, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.21484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.866455078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08447265625, "step": 513 }, { "epoch": 0.4112, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.214599609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.468505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.866455078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.083984375, "step": 514 }, { "epoch": 0.412, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.866455078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.083740234375, "step": 515 }, { "epoch": 0.4128, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8662109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.083251953125, "step": 516 }, { "epoch": 0.4136, "model.model.language_model.layers.26.mlp.down_proj": 0.435546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0830078125, "step": 517 }, { "epoch": 0.4144, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.212646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.337890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0830078125, "step": 518 }, { "epoch": 0.4152, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.212158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.804931640625, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.466552734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.082763671875, "step": 519 }, { "epoch": 0.416, "grad_norm": 0.21036136150360107, "learning_rate": 4.3066666666666665e-05, "loss": 2.36494140625, "step": 520 }, { "epoch": 0.416, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.212158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.804931640625, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.44921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.466552734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.082763671875, "step": 520 }, { "epoch": 0.4168, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.212158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.804931640625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08251953125, "step": 521 }, { "epoch": 0.4176, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4658203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.082275390625, "step": 522 }, { "epoch": 0.4184, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.211669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8037109375, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.465576171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 523 }, { "epoch": 0.4192, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.21142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8037109375, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 524 }, { "epoch": 0.42, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.211181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.464599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865966796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 525 }, { "epoch": 0.4208, "model.model.language_model.layers.26.mlp.down_proj": 0.423828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.211181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.464599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 526 }, { "epoch": 0.4216, "model.model.language_model.layers.26.mlp.down_proj": 0.439453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.211181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.464599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 527 }, { "epoch": 0.4224, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08203125, "step": 528 }, { "epoch": 0.4232, "model.model.language_model.layers.26.mlp.down_proj": 0.443359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.464111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.081787109375, "step": 529 }, { "epoch": 0.424, "grad_norm": 0.18844489753246307, "learning_rate": 4.293333333333334e-05, "loss": 2.3556640625, "step": 530 }, { "epoch": 0.424, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.210205078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08154296875, "step": 530 }, { "epoch": 0.4248, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.210205078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.081298828125, "step": 531 }, { "epoch": 0.4256, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.210205078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.081298828125, "step": 532 }, { "epoch": 0.4264, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.210205078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.803466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.081298828125, "step": 533 }, { "epoch": 0.4272, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86572265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.081298828125, "step": 534 }, { "epoch": 0.428, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.209716796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865478515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080810546875, "step": 535 }, { "epoch": 0.4288, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.209716796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.435546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865478515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080810546875, "step": 536 }, { "epoch": 0.4296, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20947265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865478515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080810546875, "step": 537 }, { "epoch": 0.4304, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20947265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44482421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.463134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.865234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080810546875, "step": 538 }, { "epoch": 0.4312, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.209228515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.444580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.462646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.08056640625, "step": 539 }, { "epoch": 0.432, "grad_norm": 0.22238987684249878, "learning_rate": 4.2800000000000004e-05, "loss": 2.30625, "step": 540 }, { "epoch": 0.432, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.208251953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.40234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.444580078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.462646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080322265625, "step": 540 }, { "epoch": 0.4328, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.208251953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 541 }, { "epoch": 0.4336, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80322265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 542 }, { "epoch": 0.4344, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802978515625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.462158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 543 }, { "epoch": 0.4352, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802978515625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.462158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 544 }, { "epoch": 0.436, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.207763671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 545 }, { "epoch": 0.4368, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.207763671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.46142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 546 }, { "epoch": 0.4376, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.207275390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.461181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 547 }, { "epoch": 0.4384, "model.model.language_model.layers.26.mlp.down_proj": 0.423828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.207275390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.461181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 548 }, { "epoch": 0.4392, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.461181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 549 }, { "epoch": 0.44, "grad_norm": 0.21074526011943817, "learning_rate": 4.266666666666667e-05, "loss": 2.3361328125, "step": 550 }, { "epoch": 0.44, "model.model.language_model.layers.26.mlp.down_proj": 0.408203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.461181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 550 }, { "epoch": 0.4408, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.206787109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.460693359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 551 }, { "epoch": 0.4416, "model.model.language_model.layers.26.mlp.down_proj": 0.3984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20654296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.460205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 552 }, { "epoch": 0.4424, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.206298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 553 }, { "epoch": 0.4432, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.205810546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864990234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 554 }, { "epoch": 0.444, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.204833984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.458740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.080078125, "step": 555 }, { "epoch": 0.4448, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20458984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.330078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.079833984375, "step": 556 }, { "epoch": 0.4456, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2041015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.079833984375, "step": 557 }, { "epoch": 0.4464, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.203857421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.079833984375, "step": 558 }, { "epoch": 0.4472, "model.model.language_model.layers.26.mlp.down_proj": 0.453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.203857421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.396484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.34375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.458984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07958984375, "step": 559 }, { "epoch": 0.448, "grad_norm": 0.18806271255016327, "learning_rate": 4.2533333333333335e-05, "loss": 2.341796875, "step": 560 }, { "epoch": 0.448, "model.model.language_model.layers.26.mlp.down_proj": 0.4453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.203857421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.408203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07958984375, "step": 560 }, { "epoch": 0.4488, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.4140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07958984375, "step": 561 }, { "epoch": 0.4496, "model.model.language_model.layers.26.mlp.down_proj": 0.4765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.203369140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.4296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.3671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.478515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4580078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07958984375, "step": 562 }, { "epoch": 0.4504, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.457763671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.079345703125, "step": 563 }, { "epoch": 0.4512, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.202880859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.079345703125, "step": 564 }, { "epoch": 0.452, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.457275390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0791015625, "step": 565 }, { "epoch": 0.4528, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80224609375, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0791015625, "step": 566 }, { "epoch": 0.4536, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.202392578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.456787109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0791015625, "step": 567 }, { "epoch": 0.4544, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.202392578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.456787109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078857421875, "step": 568 }, { "epoch": 0.4552, "model.model.language_model.layers.26.mlp.down_proj": 0.51953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2021484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.4609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.40625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.442626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.28125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.5078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45654296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07861328125, "step": 569 }, { "epoch": 0.456, "grad_norm": 0.2101268768310547, "learning_rate": 4.24e-05, "loss": 2.27685546875, "step": 570 }, { "epoch": 0.456, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20166015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.408203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.456298828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078369140625, "step": 570 }, { "epoch": 0.4568, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20166015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.455810546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078369140625, "step": 571 }, { "epoch": 0.4576, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.201171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.802001953125, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45556640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.864013671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078369140625, "step": 572 }, { "epoch": 0.4584, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.200927734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8017578125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078125, "step": 573 }, { "epoch": 0.4592, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.200927734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.8017578125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078125, "step": 574 }, { "epoch": 0.46, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.20068359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45458984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.078125, "step": 575 }, { "epoch": 0.4608, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.2001953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.441650390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.454345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077880859375, "step": 576 }, { "epoch": 0.4616, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.199951171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.453857421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077880859375, "step": 577 }, { "epoch": 0.4624, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.199462890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.453857421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077880859375, "step": 578 }, { "epoch": 0.4632, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198974609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702880859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077880859375, "step": 579 }, { "epoch": 0.464, "grad_norm": 0.185389906167984, "learning_rate": 4.226666666666667e-05, "loss": 2.27314453125, "step": 580 }, { "epoch": 0.464, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077392578125, "step": 580 }, { "epoch": 0.4648, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.453369140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86376953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.077392578125, "step": 581 }, { "epoch": 0.4656, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.452880859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076904296875, "step": 582 }, { "epoch": 0.4664, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.452880859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076904296875, "step": 583 }, { "epoch": 0.4672, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.452880859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076904296875, "step": 584 }, { "epoch": 0.468, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70263671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076904296875, "step": 585 }, { "epoch": 0.4688, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.198486328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07666015625, "step": 586 }, { "epoch": 0.4696, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45263671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07666015625, "step": 587 }, { "epoch": 0.4704, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076416015625, "step": 588 }, { "epoch": 0.4712, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076416015625, "step": 589 }, { "epoch": 0.472, "grad_norm": 0.1837451159954071, "learning_rate": 4.213333333333334e-05, "loss": 2.303125, "step": 590 }, { "epoch": 0.472, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19677734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076416015625, "step": 590 }, { "epoch": 0.4728, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.196533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.076171875, "step": 591 }, { "epoch": 0.4736, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.196044921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.075927734375, "step": 592 }, { "epoch": 0.4744, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.196044921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.075927734375, "step": 593 }, { "epoch": 0.4752, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.195556640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451904296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.075439453125, "step": 594 }, { "epoch": 0.476, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.195068359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.39453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.702392578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.075439453125, "step": 595 }, { "epoch": 0.4768, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.194580078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801513671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.074951171875, "step": 596 }, { "epoch": 0.4776, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.451171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.074951171875, "step": 597 }, { "epoch": 0.4784, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.194091796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.450927734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.863037109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07470703125, "step": 598 }, { "epoch": 0.4792, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.194091796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.450927734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86279296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07470703125, "step": 599 }, { "epoch": 0.48, "grad_norm": 0.19068150222301483, "learning_rate": 4.2e-05, "loss": 2.236328125, "step": 600 }, { "epoch": 0.48, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.194091796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.440185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45068359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86279296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.074462890625, "step": 600 }, { "epoch": 0.4808, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.193603515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45068359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86279296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.074462890625, "step": 601 }, { "epoch": 0.4816, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.193603515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.45068359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86279296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07421875, "step": 602 }, { "epoch": 0.4824, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19287109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07421875, "step": 603 }, { "epoch": 0.4832, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.192626953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.073974609375, "step": 604 }, { "epoch": 0.484, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1923828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07373046875, "step": 605 }, { "epoch": 0.4848, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.192138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07373046875, "step": 606 }, { "epoch": 0.4856, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.192138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43994140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.449951171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862548828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07373046875, "step": 607 }, { "epoch": 0.4864, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.19140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80126953125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439697265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44970703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0732421875, "step": 608 }, { "epoch": 0.4872, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.191162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439697265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.448974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.862060546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0732421875, "step": 609 }, { "epoch": 0.488, "grad_norm": 0.22399060428142548, "learning_rate": 4.186666666666667e-05, "loss": 2.2384765625, "step": 610 }, { "epoch": 0.488, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.191162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439697265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0732421875, "step": 610 }, { "epoch": 0.4888, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.190185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439697265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.448486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 611 }, { "epoch": 0.4896, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.190185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.448486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 612 }, { "epoch": 0.4904, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.190185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 613 }, { "epoch": 0.4912, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18994140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 614 }, { "epoch": 0.492, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.189697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4482421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 615 }, { "epoch": 0.4928, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.189453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.447998046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 616 }, { "epoch": 0.4936, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.447509765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 617 }, { "epoch": 0.4944, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.439453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072998046875, "step": 618 }, { "epoch": 0.4952, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1884765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.25, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07275390625, "step": 619 }, { "epoch": 0.496, "grad_norm": 0.21764162182807922, "learning_rate": 4.1733333333333336e-05, "loss": 2.2517578125, "step": 620 }, { "epoch": 0.496, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.396484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07275390625, "step": 620 }, { "epoch": 0.4968, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07275390625, "step": 621 }, { "epoch": 0.4976, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072509765625, "step": 622 }, { "epoch": 0.4984, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.187744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43896484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072509765625, "step": 623 }, { "epoch": 0.4992, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.187744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.438720703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072509765625, "step": 624 }, { "epoch": 0.5, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.187744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446533203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.861083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072021484375, "step": 625 }, { "epoch": 0.5008, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.187744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072021484375, "step": 626 }, { "epoch": 0.5016, "model.model.language_model.layers.26.mlp.down_proj": 0.3984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1875, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.072021484375, "step": 627 }, { "epoch": 0.5024, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.801025390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.071533203125, "step": 628 }, { "epoch": 0.5032, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80078125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.071533203125, "step": 629 }, { "epoch": 0.504, "grad_norm": 0.1869303584098816, "learning_rate": 4.16e-05, "loss": 2.2337890625, "step": 630 }, { "epoch": 0.504, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.186767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80078125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0712890625, "step": 630 }, { "epoch": 0.5048, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80078125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4462890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0712890625, "step": 631 }, { "epoch": 0.5056, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80078125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.446044921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.071044921875, "step": 632 }, { "epoch": 0.5064, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.186279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.80078125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44580078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.07080078125, "step": 633 }, { "epoch": 0.5072, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.800048828125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.070068359375, "step": 634 }, { "epoch": 0.508, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.185302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.800048828125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.437744140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.445556640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.070068359375, "step": 635 }, { "epoch": 0.5088, "model.model.language_model.layers.26.mlp.down_proj": 0.3984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.185302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.800048828125, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.437744140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.86083984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.34375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.070068359375, "step": 636 }, { "epoch": 0.5096, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.185302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.800048828125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.069580078125, "step": 637 }, { "epoch": 0.5104, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.185302734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7998046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.444091796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.069580078125, "step": 638 }, { "epoch": 0.5112, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.184326171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0693359375, "step": 639 }, { "epoch": 0.512, "grad_norm": 0.2000480443239212, "learning_rate": 4.146666666666667e-05, "loss": 2.19365234375, "step": 640 }, { "epoch": 0.512, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.184326171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.437255859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 640 }, { "epoch": 0.5128, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 641 }, { "epoch": 0.5136, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 642 }, { "epoch": 0.5144, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 643 }, { "epoch": 0.5152, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799560546875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 644 }, { "epoch": 0.516, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.183837890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 645 }, { "epoch": 0.5168, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.183837890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.443115234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 646 }, { "epoch": 0.5176, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44287109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06884765625, "step": 647 }, { "epoch": 0.5184, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8603515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.068603515625, "step": 648 }, { "epoch": 0.5192, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4423828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8603515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.068603515625, "step": 649 }, { "epoch": 0.52, "grad_norm": 0.17782552540302277, "learning_rate": 4.133333333333333e-05, "loss": 2.2083984375, "step": 650 }, { "epoch": 0.52, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.442138671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8603515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.068603515625, "step": 650 }, { "epoch": 0.5208, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.183349609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8603515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06787109375, "step": 651 }, { "epoch": 0.5216, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.182861328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701416015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06787109375, "step": 652 }, { "epoch": 0.5224, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.182861328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06787109375, "step": 653 }, { "epoch": 0.5232, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1826171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.860107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06787109375, "step": 654 }, { "epoch": 0.524, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.182373046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85986328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067626953125, "step": 655 }, { "epoch": 0.5248, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0673828125, "step": 656 }, { "epoch": 0.5256, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.181884765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.31640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.41796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0673828125, "step": 657 }, { "epoch": 0.5264, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.181396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0673828125, "step": 658 }, { "epoch": 0.5272, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1806640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0673828125, "step": 659 }, { "epoch": 0.528, "grad_norm": 0.19968806207180023, "learning_rate": 4.12e-05, "loss": 2.183203125, "step": 660 }, { "epoch": 0.528, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.180419921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 660 }, { "epoch": 0.5288, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.18017578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 661 }, { "epoch": 0.5296, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.179931640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 662 }, { "epoch": 0.5304, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.179931640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 663 }, { "epoch": 0.5312, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.441162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 664 }, { "epoch": 0.532, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.799072265625, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.44091796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.34375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 665 }, { "epoch": 0.5328, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.179443359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.440673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.067138671875, "step": 666 }, { "epoch": 0.5336, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.179443359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.440673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06689453125, "step": 667 }, { "epoch": 0.5344, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 668 }, { "epoch": 0.5352, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 669 }, { "epoch": 0.536, "grad_norm": 0.21344135701656342, "learning_rate": 4.106666666666667e-05, "loss": 2.21142578125, "step": 670 }, { "epoch": 0.536, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 670 }, { "epoch": 0.5368, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 671 }, { "epoch": 0.5376, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.700927734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 672 }, { "epoch": 0.5384, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798828125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.423828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 673 }, { "epoch": 0.5392, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.438720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 674 }, { "epoch": 0.54, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.436279296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.438720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066650390625, "step": 675 }, { "epoch": 0.5408, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4384765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06640625, "step": 676 }, { "epoch": 0.5416, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.438232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06640625, "step": 677 }, { "epoch": 0.5424, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178955078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.438232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 678 }, { "epoch": 0.5432, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1787109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 679 }, { "epoch": 0.544, "grad_norm": 0.17684093117713928, "learning_rate": 4.093333333333334e-05, "loss": 2.2197265625, "step": 680 }, { "epoch": 0.544, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1787109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 680 }, { "epoch": 0.5448, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 681 }, { "epoch": 0.5456, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 682 }, { "epoch": 0.5464, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.066162109375, "step": 683 }, { "epoch": 0.5472, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.437744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 684 }, { "epoch": 0.548, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.437744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 685 }, { "epoch": 0.5488, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.437744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 686 }, { "epoch": 0.5496, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.178466796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 687 }, { "epoch": 0.5504, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.177734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 688 }, { "epoch": 0.5512, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.177734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 689 }, { "epoch": 0.552, "grad_norm": 0.18510469794273376, "learning_rate": 4.08e-05, "loss": 2.19345703125, "step": 690 }, { "epoch": 0.552, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.177734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 690 }, { "epoch": 0.5528, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.177734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.437255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85888671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 691 }, { "epoch": 0.5536, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 692 }, { "epoch": 0.5544, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.177001953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06591796875, "step": 693 }, { "epoch": 0.5552, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.065673828125, "step": 694 }, { "epoch": 0.556, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.176513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.065673828125, "step": 695 }, { "epoch": 0.5568, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.176513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798583984375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.065673828125, "step": 696 }, { "epoch": 0.5576, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.176513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79833984375, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06494140625, "step": 697 }, { "epoch": 0.5584, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.176513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798095703125, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06494140625, "step": 698 }, { "epoch": 0.5592, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.176513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.798095703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.436767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06494140625, "step": 699 }, { "epoch": 0.56, "grad_norm": 0.1825551688671112, "learning_rate": 4.066666666666667e-05, "loss": 2.15546875, "step": 700 }, { "epoch": 0.56, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797607421875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 700 }, { "epoch": 0.5608, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797607421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 701 }, { "epoch": 0.5616, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797607421875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 702 }, { "epoch": 0.5624, "model.model.language_model.layers.26.mlp.down_proj": 0.4375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175537109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797607421875, "model.model.language_model.layers.26.mlp.up_proj": 0.3359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43603515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.858642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 703 }, { "epoch": 0.5632, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175537109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8583984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 704 }, { "epoch": 0.564, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175537109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8583984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 705 }, { "epoch": 0.5648, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175537109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8583984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 706 }, { "epoch": 0.5656, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17529296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.400390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8583984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064697265625, "step": 707 }, { "epoch": 0.5664, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175048828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064453125, "step": 708 }, { "epoch": 0.5672, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.175048828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.70068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064453125, "step": 709 }, { "epoch": 0.568, "grad_norm": 0.1906106173992157, "learning_rate": 4.0533333333333334e-05, "loss": 2.15546875, "step": 710 }, { "epoch": 0.568, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.700439453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064453125, "step": 710 }, { "epoch": 0.5688, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.700439453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.435302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064453125, "step": 711 }, { "epoch": 0.5696, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.7001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064453125, "step": 712 }, { "epoch": 0.5704, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699951171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85791015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 713 }, { "epoch": 0.5712, "model.model.language_model.layers.26.mlp.down_proj": 0.412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43505859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699951171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857666015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 714 }, { "epoch": 0.572, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699951171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857666015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 715 }, { "epoch": 0.5728, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699951171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857666015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 716 }, { "epoch": 0.5736, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699951171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857666015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 717 }, { "epoch": 0.5744, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857666015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 718 }, { "epoch": 0.5752, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 719 }, { "epoch": 0.576, "grad_norm": 0.18429826200008392, "learning_rate": 4.0400000000000006e-05, "loss": 2.125, "step": 720 }, { "epoch": 0.576, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174560546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 720 }, { "epoch": 0.5768, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79736328125, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 721 }, { "epoch": 0.5776, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.174072265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 722 }, { "epoch": 0.5784, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 723 }, { "epoch": 0.5792, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.064208984375, "step": 724 }, { "epoch": 0.58, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06396484375, "step": 725 }, { "epoch": 0.5808, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173583984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06396484375, "step": 726 }, { "epoch": 0.5816, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173095703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857177734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.063720703125, "step": 727 }, { "epoch": 0.5824, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173095703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69970703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857177734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0634765625, "step": 728 }, { "epoch": 0.5832, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.173095703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.699462890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.857177734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.063232421875, "step": 729 }, { "epoch": 0.584, "grad_norm": 0.18015600740909576, "learning_rate": 4.026666666666667e-05, "loss": 2.14326171875, "step": 730 }, { "epoch": 0.584, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.4296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.062744140625, "step": 730 }, { "epoch": 0.5848, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43310546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.062744140625, "step": 731 }, { "epoch": 0.5856, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.062744140625, "step": 732 }, { "epoch": 0.5864, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 733 }, { "epoch": 0.5872, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 734 }, { "epoch": 0.588, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.698974609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 735 }, { "epoch": 0.5888, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.698974609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 736 }, { "epoch": 0.5896, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 737 }, { "epoch": 0.5904, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 738 }, { "epoch": 0.5912, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 739 }, { "epoch": 0.592, "grad_norm": 0.1885373741388321, "learning_rate": 4.013333333333333e-05, "loss": 2.11689453125, "step": 740 }, { "epoch": 0.592, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1728515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856689453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0625, "step": 740 }, { "epoch": 0.5928, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.172607421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.797119140625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434814453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856689453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.062255859375, "step": 741 }, { "epoch": 0.5936, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.172119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4345703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856689453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.061767578125, "step": 742 }, { "epoch": 0.5944, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.172119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4345703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856689453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.061767578125, "step": 743 }, { "epoch": 0.5952, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8564453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.061279296875, "step": 744 }, { "epoch": 0.596, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8564453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.061279296875, "step": 745 }, { "epoch": 0.5968, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8564453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.061279296875, "step": 746 }, { "epoch": 0.5976, "model.model.language_model.layers.26.mlp.down_proj": 0.46484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.400390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.33984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.455078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8564453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 747 }, { "epoch": 0.5984, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 748 }, { "epoch": 0.5992, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 749 }, { "epoch": 0.6, "grad_norm": 0.1942032426595688, "learning_rate": 4e-05, "loss": 2.101953125, "step": 750 }, { "epoch": 0.6, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.434326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.431396484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 750 }, { "epoch": 0.6008, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796630859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 751 }, { "epoch": 0.6016, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 752 }, { "epoch": 0.6024, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 753 }, { "epoch": 0.6032, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430908203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 754 }, { "epoch": 0.604, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 755 }, { "epoch": 0.6048, "model.model.language_model.layers.26.mlp.down_proj": 0.453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.40625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.357421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.39453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 756 }, { "epoch": 0.6056, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 757 }, { "epoch": 0.6064, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 758 }, { "epoch": 0.6072, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433837890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 759 }, { "epoch": 0.608, "grad_norm": 0.1838601529598236, "learning_rate": 3.986666666666667e-05, "loss": 2.1505859375, "step": 760 }, { "epoch": 0.608, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79638671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06103515625, "step": 760 }, { "epoch": 0.6088, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.060546875, "step": 761 }, { "epoch": 0.6096, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.17138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.060546875, "step": 762 }, { "epoch": 0.6104, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.430419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.060546875, "step": 763 }, { "epoch": 0.6112, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.171142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.388671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.060546875, "step": 764 }, { "epoch": 0.612, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.060302734375, "step": 765 }, { "epoch": 0.6128, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429931640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.06005859375, "step": 766 }, { "epoch": 0.6136, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.796142578125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429931640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 767 }, { "epoch": 0.6144, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429931640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 768 }, { "epoch": 0.6152, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 769 }, { "epoch": 0.616, "grad_norm": 0.18244178593158722, "learning_rate": 3.9733333333333335e-05, "loss": 2.13408203125, "step": 770 }, { "epoch": 0.616, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69873046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 770 }, { "epoch": 0.6168, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.170654296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.698486328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 771 }, { "epoch": 0.6176, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.170654296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6982421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 772 }, { "epoch": 0.6184, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.170166015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6982421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 773 }, { "epoch": 0.6192, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.170166015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 774 }, { "epoch": 0.62, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.169921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.429443359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 775 }, { "epoch": 0.6208, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.169921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 776 }, { "epoch": 0.6216, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.169677734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.428466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 777 }, { "epoch": 0.6224, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.428466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 778 }, { "epoch": 0.6232, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.428466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 779 }, { "epoch": 0.624, "grad_norm": 0.16523422300815582, "learning_rate": 3.960000000000001e-05, "loss": 2.105859375, "step": 780 }, { "epoch": 0.624, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 780 }, { "epoch": 0.6248, "model.model.language_model.layers.26.mlp.down_proj": 0.416015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 781 }, { "epoch": 0.6256, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.433349609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 782 }, { "epoch": 0.6264, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.169189453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43310546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 783 }, { "epoch": 0.6272, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1689453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43310546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 784 }, { "epoch": 0.628, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 785 }, { "epoch": 0.6288, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 786 }, { "epoch": 0.6296, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432861328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 787 }, { "epoch": 0.6304, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 788 }, { "epoch": 0.6312, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168701171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 789 }, { "epoch": 0.632, "grad_norm": 0.20551402866840363, "learning_rate": 3.9466666666666666e-05, "loss": 2.09921875, "step": 790 }, { "epoch": 0.632, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059814453125, "step": 790 }, { "epoch": 0.6328, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4326171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.856201171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0595703125, "step": 791 }, { "epoch": 0.6336, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0595703125, "step": 792 }, { "epoch": 0.6344, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.168212890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0595703125, "step": 793 }, { "epoch": 0.6352, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 794 }, { "epoch": 0.636, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.432373046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697998046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 795 }, { "epoch": 0.6368, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 796 }, { "epoch": 0.6376, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85595703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 797 }, { "epoch": 0.6384, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 798 }, { "epoch": 0.6392, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 799 }, { "epoch": 0.64, "grad_norm": 0.1849522888660431, "learning_rate": 3.933333333333333e-05, "loss": 2.10107421875, "step": 800 }, { "epoch": 0.64, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 800 }, { "epoch": 0.6408, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 801 }, { "epoch": 0.6416, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167724609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.059326171875, "step": 802 }, { "epoch": 0.6424, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855712890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.058837890625, "step": 803 }, { "epoch": 0.6432, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.058837890625, "step": 804 }, { "epoch": 0.644, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167236328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.322265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05859375, "step": 805 }, { "epoch": 0.6448, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167236328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69775390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.058349609375, "step": 806 }, { "epoch": 0.6456, "model.model.language_model.layers.26.mlp.down_proj": 0.40625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.167236328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855224609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 807 }, { "epoch": 0.6464, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855224609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 808 }, { "epoch": 0.6472, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855224609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 809 }, { "epoch": 0.648, "grad_norm": 0.16802072525024414, "learning_rate": 3.9200000000000004e-05, "loss": 2.02890625, "step": 810 }, { "epoch": 0.648, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.166748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855224609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 810 }, { "epoch": 0.6488, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.166748046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795654296875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.855224609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 811 }, { "epoch": 0.6496, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 812 }, { "epoch": 0.6504, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 813 }, { "epoch": 0.6512, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 814 }, { "epoch": 0.652, "model.model.language_model.layers.26.mlp.down_proj": 0.466796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 815 }, { "epoch": 0.6528, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.424560546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 816 }, { "epoch": 0.6536, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 817 }, { "epoch": 0.6544, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.166259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 818 }, { "epoch": 0.6552, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.166259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 819 }, { "epoch": 0.656, "grad_norm": 0.1700490266084671, "learning_rate": 3.906666666666667e-05, "loss": 2.0326171875, "step": 820 }, { "epoch": 0.656, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.166259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431884765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85498046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 820 }, { "epoch": 0.6568, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.854736328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 821 }, { "epoch": 0.6576, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.854736328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 822 }, { "epoch": 0.6584, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057861328125, "step": 823 }, { "epoch": 0.6592, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.854248046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 824 }, { "epoch": 0.66, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.165283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.431396484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.854248046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 825 }, { "epoch": 0.6608, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.165283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.854248046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 826 }, { "epoch": 0.6616, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.165283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85400390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 827 }, { "epoch": 0.6624, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.165283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85400390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 828 }, { "epoch": 0.6632, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85400390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 829 }, { "epoch": 0.664, "grad_norm": 0.1827370971441269, "learning_rate": 3.8933333333333336e-05, "loss": 2.03955078125, "step": 830 }, { "epoch": 0.664, "model.model.language_model.layers.26.mlp.down_proj": 0.43359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.40234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.3515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85400390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 830 }, { "epoch": 0.6648, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0576171875, "step": 831 }, { "epoch": 0.6656, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697509765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057373046875, "step": 832 }, { "epoch": 0.6664, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057373046875, "step": 833 }, { "epoch": 0.6672, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.057373046875, "step": 834 }, { "epoch": 0.668, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 835 }, { "epoch": 0.6688, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 836 }, { "epoch": 0.6696, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 837 }, { "epoch": 0.6704, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1650390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43115234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 838 }, { "epoch": 0.6712, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.164794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.430908203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.697021484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 839 }, { "epoch": 0.672, "grad_norm": 0.18478752672672272, "learning_rate": 3.88e-05, "loss": 2.04423828125, "step": 840 }, { "epoch": 0.672, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.430419921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 840 }, { "epoch": 0.6728, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 841 }, { "epoch": 0.6736, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853759765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 842 }, { "epoch": 0.6744, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 843 }, { "epoch": 0.6752, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 844 }, { "epoch": 0.676, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.795166015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.853271484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 845 }, { "epoch": 0.6768, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 846 }, { "epoch": 0.6776, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 847 }, { "epoch": 0.6784, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.164306640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 848 }, { "epoch": 0.6792, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.164306640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 849 }, { "epoch": 0.68, "grad_norm": 0.17504391074180603, "learning_rate": 3.866666666666667e-05, "loss": 2.049609375, "step": 850 }, { "epoch": 0.68, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.164306640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696533203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 850 }, { "epoch": 0.6808, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05712890625, "step": 851 }, { "epoch": 0.6816, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 852 }, { "epoch": 0.6824, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.163818359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 853 }, { "epoch": 0.6832, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 854 }, { "epoch": 0.684, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 855 }, { "epoch": 0.6848, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 856 }, { "epoch": 0.6856, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 857 }, { "epoch": 0.6864, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.163330078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 858 }, { "epoch": 0.6872, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.163330078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 859 }, { "epoch": 0.688, "grad_norm": 0.1856231391429901, "learning_rate": 3.853333333333334e-05, "loss": 2.03544921875, "step": 860 }, { "epoch": 0.688, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.163330078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 860 }, { "epoch": 0.6888, "model.model.language_model.layers.26.mlp.down_proj": 0.41015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 861 }, { "epoch": 0.6896, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 862 }, { "epoch": 0.6904, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 863 }, { "epoch": 0.6912, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 864 }, { "epoch": 0.692, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794677734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.43017578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056884765625, "step": 865 }, { "epoch": 0.6928, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1630859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79443359375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85302734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 866 }, { "epoch": 0.6936, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79443359375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420654296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 867 }, { "epoch": 0.6944, "model.model.language_model.layers.26.mlp.down_proj": 0.4140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79443359375, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.3125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 868 }, { "epoch": 0.6952, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79443359375, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 869 }, { "epoch": 0.696, "grad_norm": 0.17308782041072845, "learning_rate": 3.8400000000000005e-05, "loss": 2.0455078125, "step": 870 }, { "epoch": 0.696, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.429443359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 870 }, { "epoch": 0.6968, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 871 }, { "epoch": 0.6976, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056640625, "step": 872 }, { "epoch": 0.6984, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.056396484375, "step": 873 }, { "epoch": 0.6992, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 874 }, { "epoch": 0.7, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 875 }, { "epoch": 0.7008, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6962890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 876 }, { "epoch": 0.7016, "model.model.language_model.layers.26.mlp.down_proj": 0.42578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.42578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 877 }, { "epoch": 0.7024, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 878 }, { "epoch": 0.7032, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 879 }, { "epoch": 0.704, "grad_norm": 0.1796206533908844, "learning_rate": 3.8266666666666664e-05, "loss": 2.02861328125, "step": 880 }, { "epoch": 0.704, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 880 }, { "epoch": 0.7048, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 881 }, { "epoch": 0.7056, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.696044921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05615234375, "step": 882 }, { "epoch": 0.7064, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 883 }, { "epoch": 0.7072, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428955078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 884 }, { "epoch": 0.708, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 885 }, { "epoch": 0.7088, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 886 }, { "epoch": 0.7096, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.161376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.794189453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 887 }, { "epoch": 0.7104, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160888671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 888 }, { "epoch": 0.7112, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160888671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 889 }, { "epoch": 0.712, "grad_norm": 0.18479767441749573, "learning_rate": 3.8133333333333336e-05, "loss": 2.037890625, "step": 890 }, { "epoch": 0.712, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055908203125, "step": 890 }, { "epoch": 0.7128, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 891 }, { "epoch": 0.7136, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 892 }, { "epoch": 0.7144, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 893 }, { "epoch": 0.7152, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 894 }, { "epoch": 0.716, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.160400390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7939453125, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 895 }, { "epoch": 0.7168, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852783203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 896 }, { "epoch": 0.7176, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 897 }, { "epoch": 0.7184, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0556640625, "step": 898 }, { "epoch": 0.7192, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 899 }, { "epoch": 0.72, "grad_norm": 0.1849372833967209, "learning_rate": 3.8e-05, "loss": 1.99580078125, "step": 900 }, { "epoch": 0.72, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.16015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 900 }, { "epoch": 0.7208, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 901 }, { "epoch": 0.7216, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8525390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 902 }, { "epoch": 0.7224, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 903 }, { "epoch": 0.7232, "model.model.language_model.layers.26.mlp.down_proj": 0.50390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.43359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.3671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 904 }, { "epoch": 0.724, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 905 }, { "epoch": 0.7248, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 906 }, { "epoch": 0.7256, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793701171875, "model.model.language_model.layers.26.mlp.up_proj": 0.310546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.055419921875, "step": 907 }, { "epoch": 0.7264, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695556640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 908 }, { "epoch": 0.7272, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 909 }, { "epoch": 0.728, "grad_norm": 0.1780901849269867, "learning_rate": 3.786666666666667e-05, "loss": 2.00986328125, "step": 910 }, { "epoch": 0.728, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 910 }, { "epoch": 0.7288, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4287109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.852294921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 911 }, { "epoch": 0.7296, "model.model.language_model.layers.26.mlp.down_proj": 0.4296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.428466796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 912 }, { "epoch": 0.7304, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05517578125, "step": 913 }, { "epoch": 0.7312, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159912109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42822265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 914 }, { "epoch": 0.732, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 915 }, { "epoch": 0.7328, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 916 }, { "epoch": 0.7336, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 917 }, { "epoch": 0.7344, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 918 }, { "epoch": 0.7352, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 919 }, { "epoch": 0.736, "grad_norm": 0.17590318620204926, "learning_rate": 3.773333333333334e-05, "loss": 2.01640625, "step": 920 }, { "epoch": 0.736, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851806640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 920 }, { "epoch": 0.7368, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 921 }, { "epoch": 0.7376, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 922 }, { "epoch": 0.7384, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 923 }, { "epoch": 0.7392, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.159423828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 924 }, { "epoch": 0.74, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 925 }, { "epoch": 0.7408, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 926 }, { "epoch": 0.7416, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 927 }, { "epoch": 0.7424, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 928 }, { "epoch": 0.7432, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 929 }, { "epoch": 0.744, "grad_norm": 0.15517467260360718, "learning_rate": 3.76e-05, "loss": 1.96962890625, "step": 930 }, { "epoch": 0.744, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427978515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 930 }, { "epoch": 0.7448, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054931640625, "step": 931 }, { "epoch": 0.7456, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0546875, "step": 932 }, { "epoch": 0.7464, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 933 }, { "epoch": 0.7472, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79345703125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 934 }, { "epoch": 0.748, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158935546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.43359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 935 }, { "epoch": 0.7488, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 936 }, { "epoch": 0.7496, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158447265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.793212890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 937 }, { "epoch": 0.7504, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 938 }, { "epoch": 0.7512, "model.model.language_model.layers.26.mlp.down_proj": 0.400390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 939 }, { "epoch": 0.752, "grad_norm": 0.15573440492153168, "learning_rate": 3.7466666666666665e-05, "loss": 1.99423828125, "step": 940 }, { "epoch": 0.752, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.054443359375, "step": 940 }, { "epoch": 0.7528, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.851318359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05419921875, "step": 941 }, { "epoch": 0.7536, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05419921875, "step": 942 }, { "epoch": 0.7544, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.695068359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05419921875, "step": 943 }, { "epoch": 0.7552, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05419921875, "step": 944 }, { "epoch": 0.756, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85107421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05419921875, "step": 945 }, { "epoch": 0.7568, "model.model.language_model.layers.26.mlp.down_proj": 0.451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.388671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.326171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 946 }, { "epoch": 0.7576, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 947 }, { "epoch": 0.7584, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427490234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 948 }, { "epoch": 0.7592, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79296875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 949 }, { "epoch": 0.76, "grad_norm": 0.1554388552904129, "learning_rate": 3.733333333333334e-05, "loss": 2.0015625, "step": 950 }, { "epoch": 0.76, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 950 }, { "epoch": 0.7608, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 951 }, { "epoch": 0.7616, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 952 }, { "epoch": 0.7624, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 953 }, { "epoch": 0.7632, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15771484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 954 }, { "epoch": 0.764, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.157470703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69482421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 955 }, { "epoch": 0.7648, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 956 }, { "epoch": 0.7656, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 957 }, { "epoch": 0.7664, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 958 }, { "epoch": 0.7672, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 959 }, { "epoch": 0.768, "grad_norm": 0.14942418038845062, "learning_rate": 3.72e-05, "loss": 1.9798828125, "step": 960 }, { "epoch": 0.768, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 960 }, { "epoch": 0.7688, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 961 }, { "epoch": 0.7696, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053955078125, "step": 962 }, { "epoch": 0.7704, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 963 }, { "epoch": 0.7712, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 964 }, { "epoch": 0.772, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156982421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 965 }, { "epoch": 0.7728, "model.model.language_model.layers.26.mlp.down_proj": 0.3984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 966 }, { "epoch": 0.7736, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 967 }, { "epoch": 0.7744, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694580078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 968 }, { "epoch": 0.7752, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 969 }, { "epoch": 0.776, "grad_norm": 0.1677711009979248, "learning_rate": 3.706666666666667e-05, "loss": 1.960546875, "step": 970 }, { "epoch": 0.776, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 970 }, { "epoch": 0.7768, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.850341796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 971 }, { "epoch": 0.7776, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15673828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79248046875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 972 }, { "epoch": 0.7784, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.85009765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 973 }, { "epoch": 0.7792, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 974 }, { "epoch": 0.78, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156005859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 975 }, { "epoch": 0.7808, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156005859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 976 }, { "epoch": 0.7816, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156005859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 977 }, { "epoch": 0.7824, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156005859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 978 }, { "epoch": 0.7832, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.156005859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 979 }, { "epoch": 0.784, "grad_norm": 0.16742131114006042, "learning_rate": 3.6933333333333334e-05, "loss": 1.978515625, "step": 980 }, { "epoch": 0.784, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.155517578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 980 }, { "epoch": 0.7848, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.155517578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 981 }, { "epoch": 0.7856, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.155517578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.37890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.404296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 982 }, { "epoch": 0.7864, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.155517578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 983 }, { "epoch": 0.7872, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 984 }, { "epoch": 0.788, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 985 }, { "epoch": 0.7888, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1552734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 986 }, { "epoch": 0.7896, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.155029296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 987 }, { "epoch": 0.7904, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.694091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0537109375, "step": 988 }, { "epoch": 0.7912, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.792236328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 989 }, { "epoch": 0.792, "grad_norm": 0.17740336060523987, "learning_rate": 3.68e-05, "loss": 1.9310546875, "step": 990 }, { "epoch": 0.792, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7919921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 990 }, { "epoch": 0.7928, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7919921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 991 }, { "epoch": 0.7936, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7919921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 992 }, { "epoch": 0.7944, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154541015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7919921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 993 }, { "epoch": 0.7952, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154541015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7919921875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.427001953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 994 }, { "epoch": 0.796, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4267578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 995 }, { "epoch": 0.7968, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4267578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.053466796875, "step": 996 }, { "epoch": 0.7976, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426513671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 997 }, { "epoch": 0.7984, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 998 }, { "epoch": 0.7992, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 999 }, { "epoch": 0.8, "grad_norm": 0.1772693395614624, "learning_rate": 3.6666666666666666e-05, "loss": 1.95166015625, "step": 1000 }, { "epoch": 0.8, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1000 }, { "epoch": 0.8008, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.357421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1001 }, { "epoch": 0.8016, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.41015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1002 }, { "epoch": 0.8024, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1003 }, { "epoch": 0.8032, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1004 }, { "epoch": 0.804, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1005 }, { "epoch": 0.8048, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69384765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1006 }, { "epoch": 0.8056, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1007 }, { "epoch": 0.8064, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.154052734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1008 }, { "epoch": 0.8072, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1009 }, { "epoch": 0.808, "grad_norm": 0.15788085758686066, "learning_rate": 3.653333333333334e-05, "loss": 1.9765625, "step": 1010 }, { "epoch": 0.808, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1010 }, { "epoch": 0.8088, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42626953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1011 }, { "epoch": 0.8096, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05322265625, "step": 1012 }, { "epoch": 0.8104, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1013 }, { "epoch": 0.8112, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1014 }, { "epoch": 0.812, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1015 }, { "epoch": 0.8128, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.153564453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1016 }, { "epoch": 0.8136, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.153564453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1017 }, { "epoch": 0.8144, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1018 }, { "epoch": 0.8152, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849853515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1019 }, { "epoch": 0.816, "grad_norm": 0.16713523864746094, "learning_rate": 3.6400000000000004e-05, "loss": 1.96611328125, "step": 1020 }, { "epoch": 0.816, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1020 }, { "epoch": 0.8168, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1021 }, { "epoch": 0.8176, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.426025390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1022 }, { "epoch": 0.8184, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1533203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1023 }, { "epoch": 0.8192, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.153076171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1024 }, { "epoch": 0.82, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.153076171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.693115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1025 }, { "epoch": 0.8208, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1026 }, { "epoch": 0.8216, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.416015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1027 }, { "epoch": 0.8224, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1028 }, { "epoch": 0.8232, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15283203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1029 }, { "epoch": 0.824, "grad_norm": 0.18707288801670074, "learning_rate": 3.626666666666667e-05, "loss": 1.94423828125, "step": 1030 }, { "epoch": 0.824, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1030 }, { "epoch": 0.8248, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1031 }, { "epoch": 0.8256, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1032 }, { "epoch": 0.8264, "model.model.language_model.layers.26.mlp.down_proj": 0.404296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.37109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791748046875, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1033 }, { "epoch": 0.8272, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79150390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1034 }, { "epoch": 0.828, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79150390625, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1035 }, { "epoch": 0.8288, "model.model.language_model.layers.26.mlp.down_proj": 0.423828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79150390625, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.384765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1036 }, { "epoch": 0.8296, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79150390625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1037 }, { "epoch": 0.8304, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152587890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79150390625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1038 }, { "epoch": 0.8312, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791259765625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1039 }, { "epoch": 0.832, "grad_norm": 0.15518343448638916, "learning_rate": 3.6133333333333335e-05, "loss": 1.9255859375, "step": 1040 }, { "epoch": 0.832, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791259765625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1040 }, { "epoch": 0.8328, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791259765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1041 }, { "epoch": 0.8336, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791259765625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425537109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1042 }, { "epoch": 0.8344, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791015625, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1043 }, { "epoch": 0.8352, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1044 }, { "epoch": 0.836, "model.model.language_model.layers.26.mlp.down_proj": 0.419921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791015625, "model.model.language_model.layers.26.mlp.up_proj": 0.3125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1045 }, { "epoch": 0.8368, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1046 }, { "epoch": 0.8376, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.791015625, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1047 }, { "epoch": 0.8384, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.152099609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1048 }, { "epoch": 0.8392, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1049 }, { "epoch": 0.84, "grad_norm": 0.190830260515213, "learning_rate": 3.6e-05, "loss": 1.90693359375, "step": 1050 }, { "epoch": 0.84, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1050 }, { "epoch": 0.8408, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1051 }, { "epoch": 0.8416, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42529296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1052 }, { "epoch": 0.8424, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1053 }, { "epoch": 0.8432, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1054 }, { "epoch": 0.844, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790771484375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69287109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1055 }, { "epoch": 0.8448, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.692626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1056 }, { "epoch": 0.8456, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.692626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1057 }, { "epoch": 0.8464, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15185546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1058 }, { "epoch": 0.8472, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.151611328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.425048828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1059 }, { "epoch": 0.848, "grad_norm": 0.16120034456253052, "learning_rate": 3.586666666666667e-05, "loss": 1.94287109375, "step": 1060 }, { "epoch": 0.848, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1060 }, { "epoch": 0.8488, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052978515625, "step": 1061 }, { "epoch": 0.8496, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1062 }, { "epoch": 0.8504, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1063 }, { "epoch": 0.8512, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1513671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1064 }, { "epoch": 0.852, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.151123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1065 }, { "epoch": 0.8528, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1066 }, { "epoch": 0.8536, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6923828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1067 }, { "epoch": 0.8544, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.692138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1068 }, { "epoch": 0.8552, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.15087890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.692138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1069 }, { "epoch": 0.856, "grad_norm": 0.17897270619869232, "learning_rate": 3.573333333333333e-05, "loss": 1.91162109375, "step": 1070 }, { "epoch": 0.856, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69189453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1070 }, { "epoch": 0.8568, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69189453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052734375, "step": 1071 }, { "epoch": 0.8576, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052490234375, "step": 1072 }, { "epoch": 0.8584, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1073 }, { "epoch": 0.8592, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1074 }, { "epoch": 0.86, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1075 }, { "epoch": 0.8608, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150634765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1076 }, { "epoch": 0.8616, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1077 }, { "epoch": 0.8624, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1078 }, { "epoch": 0.8632, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.79052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1079 }, { "epoch": 0.864, "grad_norm": 0.15753792226314545, "learning_rate": 3.56e-05, "loss": 1.9435546875, "step": 1080 }, { "epoch": 0.864, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790283203125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.849365234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05224609375, "step": 1080 }, { "epoch": 0.8648, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.150146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790283203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1081 }, { "epoch": 0.8656, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.790283203125, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1082 }, { "epoch": 0.8664, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1083 }, { "epoch": 0.8672, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1084 }, { "epoch": 0.868, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1085 }, { "epoch": 0.8688, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4248046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.052001953125, "step": 1086 }, { "epoch": 0.8696, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424560546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1087 }, { "epoch": 0.8704, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424560546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84912109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1088 }, { "epoch": 0.8712, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42431640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1089 }, { "epoch": 0.872, "grad_norm": 0.22808422148227692, "learning_rate": 3.546666666666667e-05, "loss": 1.90185546875, "step": 1090 }, { "epoch": 0.872, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1090 }, { "epoch": 0.8728, "model.model.language_model.layers.26.mlp.down_proj": 0.23828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1091 }, { "epoch": 0.8736, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1092 }, { "epoch": 0.8744, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.691162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1093 }, { "epoch": 0.8752, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1494140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1094 }, { "epoch": 0.876, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.1826171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.424072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.69091796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1095 }, { "epoch": 0.8768, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1096 }, { "epoch": 0.8776, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1097 }, { "epoch": 0.8784, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1098 }, { "epoch": 0.8792, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.369140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1099 }, { "epoch": 0.88, "grad_norm": 0.16375099122524261, "learning_rate": 3.5333333333333336e-05, "loss": 1.91748046875, "step": 1100 }, { "epoch": 0.88, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1100 }, { "epoch": 0.8808, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848876953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1101 }, { "epoch": 0.8816, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8486328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1102 }, { "epoch": 0.8824, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8486328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1103 }, { "epoch": 0.8832, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6904296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8486328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1104 }, { "epoch": 0.884, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6904296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1105 }, { "epoch": 0.8848, "model.model.language_model.layers.26.mlp.down_proj": 0.431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148681640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.380859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6904296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.419921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1106 }, { "epoch": 0.8856, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423583984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4033203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1107 }, { "epoch": 0.8864, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4033203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1108 }, { "epoch": 0.8872, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4033203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1109 }, { "epoch": 0.888, "grad_norm": 0.15959303081035614, "learning_rate": 3.52e-05, "loss": 1.93291015625, "step": 1110 }, { "epoch": 0.888, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0517578125, "step": 1110 }, { "epoch": 0.8888, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.36328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1111 }, { "epoch": 0.8896, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1112 }, { "epoch": 0.8904, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1113 }, { "epoch": 0.8912, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1114 }, { "epoch": 0.892, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1115 }, { "epoch": 0.8928, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1116 }, { "epoch": 0.8936, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148193359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1117 }, { "epoch": 0.8944, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148193359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42333984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1118 }, { "epoch": 0.8952, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.148193359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7900390625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1119 }, { "epoch": 0.896, "grad_norm": 0.1794893741607666, "learning_rate": 3.506666666666667e-05, "loss": 1.93515625, "step": 1120 }, { "epoch": 0.896, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1120 }, { "epoch": 0.8968, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1121 }, { "epoch": 0.8976, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789794921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.690185546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1122 }, { "epoch": 0.8984, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68994140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1123 }, { "epoch": 0.8992, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14794921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1124 }, { "epoch": 0.9, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051513671875, "step": 1125 }, { "epoch": 0.9008, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1126 }, { "epoch": 0.9016, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1127 }, { "epoch": 0.9024, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1128 }, { "epoch": 0.9032, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78955078125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689697265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1129 }, { "epoch": 0.904, "grad_norm": 0.16189126670360565, "learning_rate": 3.493333333333333e-05, "loss": 1.923828125, "step": 1130 }, { "epoch": 0.904, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1130 }, { "epoch": 0.9048, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1131 }, { "epoch": 0.9056, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.349609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1132 }, { "epoch": 0.9064, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1133 }, { "epoch": 0.9072, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1134 }, { "epoch": 0.908, "model.model.language_model.layers.26.mlp.down_proj": 0.388671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147705078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1135 }, { "epoch": 0.9088, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1136 }, { "epoch": 0.9096, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.147216796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1137 }, { "epoch": 0.9104, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1138 }, { "epoch": 0.9112, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1139 }, { "epoch": 0.912, "grad_norm": 0.18924583494663239, "learning_rate": 3.48e-05, "loss": 1.9208984375, "step": 1140 }, { "epoch": 0.912, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1140 }, { "epoch": 0.9128, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.423095703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1141 }, { "epoch": 0.9136, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1142 }, { "epoch": 0.9144, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14697265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.848388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1143 }, { "epoch": 0.9152, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05126953125, "step": 1144 }, { "epoch": 0.916, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1145 }, { "epoch": 0.9168, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.789306640625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1146 }, { "epoch": 0.9176, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1147 }, { "epoch": 0.9184, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1148 }, { "epoch": 0.9192, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1149 }, { "epoch": 0.92, "grad_norm": 0.16383355855941772, "learning_rate": 3.466666666666667e-05, "loss": 1.89482421875, "step": 1150 }, { "epoch": 0.92, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.051025390625, "step": 1150 }, { "epoch": 0.9208, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1151 }, { "epoch": 0.9216, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1152 }, { "epoch": 0.9224, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1153 }, { "epoch": 0.9232, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7890625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1154 }, { "epoch": 0.924, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788818359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84814453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1155 }, { "epoch": 0.9248, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788818359375, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1156 }, { "epoch": 0.9256, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788818359375, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1157 }, { "epoch": 0.9264, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788818359375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1158 }, { "epoch": 0.9272, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4228515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68896484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1159 }, { "epoch": 0.928, "grad_norm": 0.17893777787685394, "learning_rate": 3.453333333333334e-05, "loss": 1.88798828125, "step": 1160 }, { "epoch": 0.928, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688720703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1160 }, { "epoch": 0.9288, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688720703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05078125, "step": 1161 }, { "epoch": 0.9296, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.400634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1162 }, { "epoch": 0.9304, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.400634765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1163 }, { "epoch": 0.9312, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.146240234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1164 }, { "epoch": 0.932, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1455078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422607421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.400146484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1165 }, { "epoch": 0.9328, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1166 }, { "epoch": 0.9336, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1167 }, { "epoch": 0.9344, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1168 }, { "epoch": 0.9352, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1169 }, { "epoch": 0.936, "grad_norm": 0.16790296137332916, "learning_rate": 3.4399999999999996e-05, "loss": 1.862109375, "step": 1170 }, { "epoch": 0.936, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.399658203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050537109375, "step": 1170 }, { "epoch": 0.9368, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.318359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.399658203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1171 }, { "epoch": 0.9376, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.399658203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1172 }, { "epoch": 0.9384, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3994140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1173 }, { "epoch": 0.9392, "model.model.language_model.layers.26.mlp.down_proj": 0.421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.38671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.32421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.399169921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1174 }, { "epoch": 0.94, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.422119140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.399169921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1175 }, { "epoch": 0.9408, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.05029296875, "step": 1176 }, { "epoch": 0.9416, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050048828125, "step": 1177 }, { "epoch": 0.9424, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050048828125, "step": 1178 }, { "epoch": 0.9432, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050048828125, "step": 1179 }, { "epoch": 0.944, "grad_norm": 0.18652097880840302, "learning_rate": 3.426666666666667e-05, "loss": 1.91123046875, "step": 1180 }, { "epoch": 0.944, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.050048828125, "step": 1180 }, { "epoch": 0.9448, "model.model.language_model.layers.26.mlp.down_proj": 0.3984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.145263671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.298828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.398681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1181 }, { "epoch": 0.9456, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.398681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1182 }, { "epoch": 0.9464, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1183 }, { "epoch": 0.9472, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1184 }, { "epoch": 0.948, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1185 }, { "epoch": 0.9488, "model.model.language_model.layers.26.mlp.down_proj": 0.435546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.40625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1186 }, { "epoch": 0.9496, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1187 }, { "epoch": 0.9504, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1188 }, { "epoch": 0.9512, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1189 }, { "epoch": 0.952, "grad_norm": 0.17044875025749207, "learning_rate": 3.4133333333333334e-05, "loss": 1.90849609375, "step": 1190 }, { "epoch": 0.952, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1190 }, { "epoch": 0.9528, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1191 }, { "epoch": 0.9536, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1192 }, { "epoch": 0.9544, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1193 }, { "epoch": 0.9552, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1194 }, { "epoch": 0.956, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1195 }, { "epoch": 0.9568, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.398193359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1196 }, { "epoch": 0.9576, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1197 }, { "epoch": 0.9584, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144775390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1198 }, { "epoch": 0.9592, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1199 }, { "epoch": 0.96, "grad_norm": 0.17590853571891785, "learning_rate": 3.4000000000000007e-05, "loss": 1.87646484375, "step": 1200 }, { "epoch": 0.96, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847900390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1200 }, { "epoch": 0.9608, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39794921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1201 }, { "epoch": 0.9616, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.36328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397705078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1202 }, { "epoch": 0.9624, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397705078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1203 }, { "epoch": 0.9632, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.144287109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.688232421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397705078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1204 }, { "epoch": 0.964, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397705078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1205 }, { "epoch": 0.9648, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1206 }, { "epoch": 0.9656, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1207 }, { "epoch": 0.9664, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421630859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1208 }, { "epoch": 0.9672, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1209 }, { "epoch": 0.968, "grad_norm": 0.17394159734249115, "learning_rate": 3.3866666666666665e-05, "loss": 1.89365234375, "step": 1210 }, { "epoch": 0.968, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1210 }, { "epoch": 0.9688, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1211 }, { "epoch": 0.9696, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78857421875, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68798828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3974609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1212 }, { "epoch": 0.9704, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687744140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1213 }, { "epoch": 0.9712, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687744140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1214 }, { "epoch": 0.972, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687744140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.397216796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1215 }, { "epoch": 0.9728, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687744140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1216 }, { "epoch": 0.9736, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687744140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1217 }, { "epoch": 0.9744, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1218 }, { "epoch": 0.9752, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1219 }, { "epoch": 0.976, "grad_norm": 0.1563493013381958, "learning_rate": 3.373333333333333e-05, "loss": 1.8603515625, "step": 1220 }, { "epoch": 0.976, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687255859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1220 }, { "epoch": 0.9768, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687255859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1221 }, { "epoch": 0.9776, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687255859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1222 }, { "epoch": 0.9784, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.341796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.687255859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1223 }, { "epoch": 0.9792, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1224 }, { "epoch": 0.98, "model.model.language_model.layers.26.mlp.down_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68701171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1225 }, { "epoch": 0.9808, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1226 }, { "epoch": 0.9816, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1227 }, { "epoch": 0.9824, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.788330078125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1228 }, { "epoch": 0.9832, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1229 }, { "epoch": 0.984, "grad_norm": 0.1555301994085312, "learning_rate": 3.3600000000000004e-05, "loss": 1.8873046875, "step": 1230 }, { "epoch": 0.984, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.847412109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1230 }, { "epoch": 0.9848, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1231 }, { "epoch": 0.9856, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39697265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1232 }, { "epoch": 0.9864, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42138671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0498046875, "step": 1233 }, { "epoch": 0.9872, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049560546875, "step": 1234 }, { "epoch": 0.988, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049560546875, "step": 1235 }, { "epoch": 0.9888, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049560546875, "step": 1236 }, { "epoch": 0.9896, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049560546875, "step": 1237 }, { "epoch": 0.9904, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1238 }, { "epoch": 0.9912, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1239 }, { "epoch": 0.992, "grad_norm": 0.1830337941646576, "learning_rate": 3.346666666666667e-05, "loss": 1.8765625, "step": 1240 }, { "epoch": 0.992, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1240 }, { "epoch": 0.9928, "model.model.language_model.layers.26.mlp.down_proj": 0.427734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.3203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.380859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1241 }, { "epoch": 0.9936, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.421142578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1242 }, { "epoch": 0.9944, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1243 }, { "epoch": 0.9952, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1244 }, { "epoch": 0.996, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1245 }, { "epoch": 0.9968, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1246 }, { "epoch": 0.9976, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4208984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1247 }, { "epoch": 0.9984, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420654296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1248 }, { "epoch": 0.9992, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.37109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1249 }, { "epoch": 1.0, "grad_norm": 0.16670732200145721, "learning_rate": 3.3333333333333335e-05, "loss": 1.87099609375, "step": 1250 }, { "epoch": 1.0, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1250 }, { "epoch": 1.0008, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1251 }, { "epoch": 1.0016, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1252 }, { "epoch": 1.0024, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396728515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1253 }, { "epoch": 1.0032, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1254 }, { "epoch": 1.004, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1255 }, { "epoch": 1.0048, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1256 }, { "epoch": 1.0056, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.686279296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.396240234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1257 }, { "epoch": 1.0064, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1258 }, { "epoch": 1.0072, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39599609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1259 }, { "epoch": 1.008, "grad_norm": 0.1647653728723526, "learning_rate": 3.32e-05, "loss": 1.88408203125, "step": 1260 }, { "epoch": 1.008, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1260 }, { "epoch": 1.0088, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1261 }, { "epoch": 1.0096, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1262 }, { "epoch": 1.0104, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1263 }, { "epoch": 1.0112, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1264 }, { "epoch": 1.012, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1265 }, { "epoch": 1.0128, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.143798828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.395751953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1266 }, { "epoch": 1.0136, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3955078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1267 }, { "epoch": 1.0144, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3955078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1268 }, { "epoch": 1.0152, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04931640625, "step": 1269 }, { "epoch": 1.016, "grad_norm": 0.15146368741989136, "learning_rate": 3.3066666666666666e-05, "loss": 1.86005859375, "step": 1270 }, { "epoch": 1.016, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1270 }, { "epoch": 1.0168, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1271 }, { "epoch": 1.0176, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1272 }, { "epoch": 1.0184, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1273 }, { "epoch": 1.0192, "model.model.language_model.layers.26.mlp.down_proj": 0.41796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.361328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.30859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846923828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1274 }, { "epoch": 1.02, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1275 }, { "epoch": 1.0208, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1276 }, { "epoch": 1.0216, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1277 }, { "epoch": 1.0224, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1278 }, { "epoch": 1.0232, "model.model.language_model.layers.26.mlp.down_proj": 0.39453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1279 }, { "epoch": 1.024, "grad_norm": 0.1705479621887207, "learning_rate": 3.293333333333333e-05, "loss": 1.83642578125, "step": 1280 }, { "epoch": 1.024, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1280 }, { "epoch": 1.0248, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.394775390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1281 }, { "epoch": 1.0256, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1282 }, { "epoch": 1.0264, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1283 }, { "epoch": 1.0272, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1284 }, { "epoch": 1.028, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78759765625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8466796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1285 }, { "epoch": 1.0288, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.394287109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1286 }, { "epoch": 1.0296, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.394287109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1287 }, { "epoch": 1.0304, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.394287109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1288 }, { "epoch": 1.0312, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1289 }, { "epoch": 1.032, "grad_norm": 0.1657629907131195, "learning_rate": 3.2800000000000004e-05, "loss": 1.84462890625, "step": 1290 }, { "epoch": 1.032, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1290 }, { "epoch": 1.0328, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1291 }, { "epoch": 1.0336, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1292 }, { "epoch": 1.0344, "model.model.language_model.layers.26.mlp.down_proj": 0.244140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142822265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1293 }, { "epoch": 1.0352, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3935546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1294 }, { "epoch": 1.036, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3935546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1295 }, { "epoch": 1.0368, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1296 }, { "epoch": 1.0376, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1297 }, { "epoch": 1.0384, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1298 }, { "epoch": 1.0392, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1299 }, { "epoch": 1.04, "grad_norm": 0.15891651809215546, "learning_rate": 3.266666666666667e-05, "loss": 1.85517578125, "step": 1300 }, { "epoch": 1.04, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1300 }, { "epoch": 1.0408, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39306640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1301 }, { "epoch": 1.0416, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.787109375, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68603515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392822265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1302 }, { "epoch": 1.0424, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.786865234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1303 }, { "epoch": 1.0432, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78662109375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1304 }, { "epoch": 1.044, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78662109375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1305 }, { "epoch": 1.0448, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1306 }, { "epoch": 1.0456, "model.model.language_model.layers.26.mlp.down_proj": 0.396484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.376953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1307 }, { "epoch": 1.0464, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1308 }, { "epoch": 1.0472, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1309 }, { "epoch": 1.048, "grad_norm": 0.170572891831398, "learning_rate": 3.253333333333333e-05, "loss": 1.855078125, "step": 1310 }, { "epoch": 1.048, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685791015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1310 }, { "epoch": 1.0488, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.392333984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1311 }, { "epoch": 1.0496, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1312 }, { "epoch": 1.0504, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39208984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1313 }, { "epoch": 1.0512, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.142333984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7861328125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.685546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.391845703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1314 }, { "epoch": 1.052, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14208984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.391845703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1315 }, { "epoch": 1.0528, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141845703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.391357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.846435546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1316 }, { "epoch": 1.0536, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141845703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1317 }, { "epoch": 1.0544, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141845703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39111328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1318 }, { "epoch": 1.0552, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390869140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1319 }, { "epoch": 1.056, "grad_norm": 0.18571752309799194, "learning_rate": 3.24e-05, "loss": 1.85244140625, "step": 1320 }, { "epoch": 1.056, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390869140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1320 }, { "epoch": 1.0568, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390869140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1321 }, { "epoch": 1.0576, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390869140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1322 }, { "epoch": 1.0584, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.42041015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1323 }, { "epoch": 1.0592, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1324 }, { "epoch": 1.06, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1325 }, { "epoch": 1.0608, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1326 }, { "epoch": 1.0616, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1327 }, { "epoch": 1.0624, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1328 }, { "epoch": 1.0632, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1329 }, { "epoch": 1.064, "grad_norm": 0.1949768364429474, "learning_rate": 3.226666666666667e-05, "loss": 1.82060546875, "step": 1330 }, { "epoch": 1.064, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1330 }, { "epoch": 1.0648, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1416015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785888671875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.390380859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1331 }, { "epoch": 1.0656, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1332 }, { "epoch": 1.0664, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1333 }, { "epoch": 1.0672, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1334 }, { "epoch": 1.068, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1335 }, { "epoch": 1.0688, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1336 }, { "epoch": 1.0695999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1337 }, { "epoch": 1.0704, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1338 }, { "epoch": 1.0712, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1339 }, { "epoch": 1.072, "grad_norm": 0.15717419981956482, "learning_rate": 3.213333333333334e-05, "loss": 1.82138671875, "step": 1340 }, { "epoch": 1.072, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1340 }, { "epoch": 1.0728, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1341 }, { "epoch": 1.0735999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68505859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1342 }, { "epoch": 1.0744, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1343 }, { "epoch": 1.0752, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.39013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1344 }, { "epoch": 1.076, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.141357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.389892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1345 }, { "epoch": 1.0768, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.389892578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1346 }, { "epoch": 1.0776, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.420166015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.389404296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84619140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1347 }, { "epoch": 1.0784, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1348 }, { "epoch": 1.0792, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1349 }, { "epoch": 1.08, "grad_norm": 0.1545376479625702, "learning_rate": 3.2000000000000005e-05, "loss": 1.859375, "step": 1350 }, { "epoch": 1.08, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684814453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1350 }, { "epoch": 1.0808, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1351 }, { "epoch": 1.0816, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1352 }, { "epoch": 1.0824, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1353 }, { "epoch": 1.0832, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388916015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1354 }, { "epoch": 1.084, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1355 }, { "epoch": 1.0848, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1356 }, { "epoch": 1.0856, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1357 }, { "epoch": 1.0864, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1358 }, { "epoch": 1.0872, "model.model.language_model.layers.26.mlp.down_proj": 0.23046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1359 }, { "epoch": 1.088, "grad_norm": 0.18574564158916473, "learning_rate": 3.1866666666666664e-05, "loss": 1.83935546875, "step": 1360 }, { "epoch": 1.088, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1360 }, { "epoch": 1.0888, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1361 }, { "epoch": 1.0896, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1362 }, { "epoch": 1.0904, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1363 }, { "epoch": 1.0912, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1364 }, { "epoch": 1.092, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1365 }, { "epoch": 1.0928, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1366 }, { "epoch": 1.0936, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1367 }, { "epoch": 1.0944, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1368 }, { "epoch": 1.0952, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14111328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1369 }, { "epoch": 1.096, "grad_norm": 0.14943397045135498, "learning_rate": 3.173333333333334e-05, "loss": 1.83935546875, "step": 1370 }, { "epoch": 1.096, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1370 }, { "epoch": 1.0968, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785400390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.388427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.049072265625, "step": 1371 }, { "epoch": 1.0976, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.785400390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048828125, "step": 1372 }, { "epoch": 1.0984, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048828125, "step": 1373 }, { "epoch": 1.0992, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048828125, "step": 1374 }, { "epoch": 1.1, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1375 }, { "epoch": 1.1008, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1376 }, { "epoch": 1.1016, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1377 }, { "epoch": 1.1024, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140869140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1378 }, { "epoch": 1.1032, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1379 }, { "epoch": 1.104, "grad_norm": 0.1577000916004181, "learning_rate": 3.16e-05, "loss": 1.829296875, "step": 1380 }, { "epoch": 1.104, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1380 }, { "epoch": 1.1048, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1381 }, { "epoch": 1.1056, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1382 }, { "epoch": 1.1064, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.387939453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1383 }, { "epoch": 1.1072, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.387939453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1384 }, { "epoch": 1.108, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.357421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1385 }, { "epoch": 1.1088, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.140380859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1386 }, { "epoch": 1.1096, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1387 }, { "epoch": 1.1104, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1388 }, { "epoch": 1.1112, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1389 }, { "epoch": 1.112, "grad_norm": 0.15173140168190002, "learning_rate": 3.146666666666667e-05, "loss": 1.82509765625, "step": 1390 }, { "epoch": 1.112, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1390 }, { "epoch": 1.1128, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845947265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1391 }, { "epoch": 1.1136, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419677734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1392 }, { "epoch": 1.1144, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.14013671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1393 }, { "epoch": 1.1152, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1394 }, { "epoch": 1.116, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1395 }, { "epoch": 1.1168, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1396 }, { "epoch": 1.1176, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1397 }, { "epoch": 1.1184, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1398 }, { "epoch": 1.1192, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41943359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1399 }, { "epoch": 1.12, "grad_norm": 0.1547883003950119, "learning_rate": 3.1333333333333334e-05, "loss": 1.80654296875, "step": 1400 }, { "epoch": 1.12, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.387451171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1400 }, { "epoch": 1.1208, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.684326171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1401 }, { "epoch": 1.1216, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1402 }, { "epoch": 1.1224, "model.model.language_model.layers.26.mlp.down_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.1767578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1403 }, { "epoch": 1.1232, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1404 }, { "epoch": 1.124, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.419189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1405 }, { "epoch": 1.1248, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.386962890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1406 }, { "epoch": 1.1256, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4189453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.386962890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1407 }, { "epoch": 1.1264, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1408 }, { "epoch": 1.1272, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1409 }, { "epoch": 1.1280000000000001, "grad_norm": 0.17525339126586914, "learning_rate": 3.12e-05, "loss": 1.81201171875, "step": 1410 }, { "epoch": 1.1280000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1410 }, { "epoch": 1.1288, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1411 }, { "epoch": 1.1296, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1412 }, { "epoch": 1.1304, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1413 }, { "epoch": 1.1312, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1414 }, { "epoch": 1.1320000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1415 }, { "epoch": 1.1328, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048583984375, "step": 1416 }, { "epoch": 1.1336, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1417 }, { "epoch": 1.1344, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1418 }, { "epoch": 1.1352, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1419 }, { "epoch": 1.1360000000000001, "grad_norm": 0.14954997599124908, "learning_rate": 3.1066666666666665e-05, "loss": 1.79599609375, "step": 1420 }, { "epoch": 1.1360000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1420 }, { "epoch": 1.1368, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139892578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1421 }, { "epoch": 1.1376, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1422 }, { "epoch": 1.1384, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1423 }, { "epoch": 1.1392, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1424 }, { "epoch": 1.1400000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418701171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.845458984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1425 }, { "epoch": 1.1408, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385498046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1426 }, { "epoch": 1.1416, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.385009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1427 }, { "epoch": 1.1424, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.384521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1428 }, { "epoch": 1.1432, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784912109375, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.384521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1429 }, { "epoch": 1.144, "grad_norm": 0.15201568603515625, "learning_rate": 3.093333333333334e-05, "loss": 1.80439453125, "step": 1430 }, { "epoch": 1.144, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.384521484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1430 }, { "epoch": 1.1448, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1431 }, { "epoch": 1.1456, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1432 }, { "epoch": 1.1464, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1433 }, { "epoch": 1.1472, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1434 }, { "epoch": 1.148, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38427734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1435 }, { "epoch": 1.1488, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1396484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1436 }, { "epoch": 1.1496, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.139404296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1437 }, { "epoch": 1.1504, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1438 }, { "epoch": 1.1512, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1439 }, { "epoch": 1.152, "grad_norm": 0.15319889783859253, "learning_rate": 3.08e-05, "loss": 1.7908203125, "step": 1440 }, { "epoch": 1.152, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3837890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1440 }, { "epoch": 1.1528, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1441 }, { "epoch": 1.1536, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84521484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1442 }, { "epoch": 1.1544, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1443 }, { "epoch": 1.1552, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68408203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1444 }, { "epoch": 1.156, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683837890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1445 }, { "epoch": 1.1568, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41845703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683837890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1446 }, { "epoch": 1.1576, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683837890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1447 }, { "epoch": 1.1584, "model.model.language_model.layers.26.mlp.down_proj": 0.234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683837890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1448 }, { "epoch": 1.1592, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1449 }, { "epoch": 1.16, "grad_norm": 0.16394200921058655, "learning_rate": 3.066666666666667e-05, "loss": 1.7896484375, "step": 1450 }, { "epoch": 1.16, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1450 }, { "epoch": 1.1608, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1451 }, { "epoch": 1.1616, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1452 }, { "epoch": 1.1623999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1453 }, { "epoch": 1.1632, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383544921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1454 }, { "epoch": 1.164, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1455 }, { "epoch": 1.1648, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1456 }, { "epoch": 1.1656, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38330078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1457 }, { "epoch": 1.1663999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383056640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1458 }, { "epoch": 1.1672, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.683349609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.383056640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1459 }, { "epoch": 1.168, "grad_norm": 0.1485176831483841, "learning_rate": 3.0533333333333335e-05, "loss": 1.78974609375, "step": 1460 }, { "epoch": 1.168, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138916015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1460 }, { "epoch": 1.1688, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1461 }, { "epoch": 1.1696, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382568359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04833984375, "step": 1462 }, { "epoch": 1.1703999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.365234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382568359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1463 }, { "epoch": 1.1712, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382568359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1464 }, { "epoch": 1.172, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1465 }, { "epoch": 1.1728, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1466 }, { "epoch": 1.1736, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1467 }, { "epoch": 1.1743999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1468 }, { "epoch": 1.1752, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1469 }, { "epoch": 1.176, "grad_norm": 0.1795387715101242, "learning_rate": 3.04e-05, "loss": 1.80390625, "step": 1470 }, { "epoch": 1.176, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1470 }, { "epoch": 1.1768, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1471 }, { "epoch": 1.1776, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68310546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1472 }, { "epoch": 1.1784, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844970703125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1473 }, { "epoch": 1.1792, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.418212890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1474 }, { "epoch": 1.18, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38232421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1475 }, { "epoch": 1.1808, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.048095703125, "step": 1476 }, { "epoch": 1.1816, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1477 }, { "epoch": 1.1824, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1478 }, { "epoch": 1.1832, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1479 }, { "epoch": 1.184, "grad_norm": 0.20406654477119446, "learning_rate": 3.0266666666666666e-05, "loss": 1.8130859375, "step": 1480 }, { "epoch": 1.184, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.138427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1480 }, { "epoch": 1.1848, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13818359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682861328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1481 }, { "epoch": 1.1856, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13818359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.382080078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1482 }, { "epoch": 1.1864, "model.model.language_model.layers.26.mlp.down_proj": 0.390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1483 }, { "epoch": 1.1872, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1484 }, { "epoch": 1.188, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1485 }, { "epoch": 1.1888, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1486 }, { "epoch": 1.1896, "model.model.language_model.layers.26.mlp.down_proj": 0.21875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1487 }, { "epoch": 1.1904, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1488 }, { "epoch": 1.1912, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1489 }, { "epoch": 1.192, "grad_norm": 0.20929858088493347, "learning_rate": 3.0133333333333335e-05, "loss": 1.78544921875, "step": 1490 }, { "epoch": 1.192, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1490 }, { "epoch": 1.1928, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1491 }, { "epoch": 1.1936, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417724609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6826171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1492 }, { "epoch": 1.1944, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1493 }, { "epoch": 1.1952, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1494 }, { "epoch": 1.196, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3818359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1495 }, { "epoch": 1.1968, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1496 }, { "epoch": 1.1976, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.365234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.306640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.844482421875, "model.model.language_model.layers.26.self_attn.v_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1497 }, { "epoch": 1.1984, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1498 }, { "epoch": 1.1992, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1499 }, { "epoch": 1.2, "grad_norm": 0.16487441956996918, "learning_rate": 3e-05, "loss": 1.7974609375, "step": 1500 }, { "epoch": 1.2, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.682373046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1500 }, { "epoch": 1.2008, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1501 }, { "epoch": 1.2016, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1502 }, { "epoch": 1.2024, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78466796875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1503 }, { "epoch": 1.2032, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0478515625, "step": 1504 }, { "epoch": 1.204, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047607421875, "step": 1505 }, { "epoch": 1.2048, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047607421875, "step": 1506 }, { "epoch": 1.2056, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047607421875, "step": 1507 }, { "epoch": 1.2064, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1508 }, { "epoch": 1.2072, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381591796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1509 }, { "epoch": 1.208, "grad_norm": 0.15117377042770386, "learning_rate": 2.986666666666667e-05, "loss": 1.8068359375, "step": 1510 }, { "epoch": 1.208, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1510 }, { "epoch": 1.2088, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.3046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.417236328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1511 }, { "epoch": 1.2096, "model.model.language_model.layers.26.mlp.down_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38134765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1512 }, { "epoch": 1.2104, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1513 }, { "epoch": 1.2112, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1514 }, { "epoch": 1.212, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1376953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1515 }, { "epoch": 1.2128, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1516 }, { "epoch": 1.2136, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1517 }, { "epoch": 1.2144, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1518 }, { "epoch": 1.2152, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1519 }, { "epoch": 1.216, "grad_norm": 0.16262619197368622, "learning_rate": 2.9733333333333336e-05, "loss": 1.81845703125, "step": 1520 }, { "epoch": 1.216, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4169921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1520 }, { "epoch": 1.2168, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1521 }, { "epoch": 1.2176, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1522 }, { "epoch": 1.2184, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1523 }, { "epoch": 1.2192, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1524 }, { "epoch": 1.22, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.784423828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1525 }, { "epoch": 1.2208, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1526 }, { "epoch": 1.2216, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1527 }, { "epoch": 1.2224, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1528 }, { "epoch": 1.2232, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.1767578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1529 }, { "epoch": 1.224, "grad_norm": 0.14928525686264038, "learning_rate": 2.96e-05, "loss": 1.797265625, "step": 1530 }, { "epoch": 1.224, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7841796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1530 }, { "epoch": 1.2248, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783935546875, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1531 }, { "epoch": 1.2256, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681884765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1532 }, { "epoch": 1.2264, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416748046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1533 }, { "epoch": 1.2272, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1534 }, { "epoch": 1.228, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.381103515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1535 }, { "epoch": 1.2288000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.380615234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1536 }, { "epoch": 1.2296, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41650390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.380615234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1537 }, { "epoch": 1.2304, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.380615234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1538 }, { "epoch": 1.2312, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1539 }, { "epoch": 1.232, "grad_norm": 0.1707266867160797, "learning_rate": 2.946666666666667e-05, "loss": 1.80146484375, "step": 1540 }, { "epoch": 1.232, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.38037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1540 }, { "epoch": 1.2328000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1541 }, { "epoch": 1.2336, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1542 }, { "epoch": 1.2344, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681396484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1543 }, { "epoch": 1.2352, "model.model.language_model.layers.26.mlp.down_proj": 0.216796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.1845703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681396484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1544 }, { "epoch": 1.236, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681396484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1545 }, { "epoch": 1.2368000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681396484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1546 }, { "epoch": 1.2376, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.681396484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3798828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1547 }, { "epoch": 1.2384, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379638671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1548 }, { "epoch": 1.2392, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68115234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379638671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1549 }, { "epoch": 1.24, "grad_norm": 0.1484082192182541, "learning_rate": 2.9333333333333336e-05, "loss": 1.758984375, "step": 1550 }, { "epoch": 1.24, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37939453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1550 }, { "epoch": 1.2408, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1551 }, { "epoch": 1.2416, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78369140625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1552 }, { "epoch": 1.2424, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783447265625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1553 }, { "epoch": 1.2432, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783447265625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1554 }, { "epoch": 1.244, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783447265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1555 }, { "epoch": 1.2448, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84423828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1556 }, { "epoch": 1.2456, "model.model.language_model.layers.26.mlp.down_proj": 0.359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1557 }, { "epoch": 1.2464, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1558 }, { "epoch": 1.2472, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1559 }, { "epoch": 1.248, "grad_norm": 0.1556800752878189, "learning_rate": 2.9199999999999998e-05, "loss": 1.79423828125, "step": 1560 }, { "epoch": 1.248, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1560 }, { "epoch": 1.2488, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.379150390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1561 }, { "epoch": 1.2496, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1562 }, { "epoch": 1.2504, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1563 }, { "epoch": 1.2511999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1564 }, { "epoch": 1.252, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1565 }, { "epoch": 1.2528000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1566 }, { "epoch": 1.2536, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1567 }, { "epoch": 1.2544, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1568 }, { "epoch": 1.2551999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1569 }, { "epoch": 1.256, "grad_norm": 0.17570282518863678, "learning_rate": 2.906666666666667e-05, "loss": 1.759765625, "step": 1570 }, { "epoch": 1.256, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1570 }, { "epoch": 1.2568, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1571 }, { "epoch": 1.2576, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.137451171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1572 }, { "epoch": 1.2584, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843994140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1573 }, { "epoch": 1.2591999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1574 }, { "epoch": 1.26, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1575 }, { "epoch": 1.2608, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04736328125, "step": 1576 }, { "epoch": 1.2616, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1577 }, { "epoch": 1.2624, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1578 }, { "epoch": 1.2631999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.680908203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37841796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1579 }, { "epoch": 1.264, "grad_norm": 0.14703886210918427, "learning_rate": 2.8933333333333333e-05, "loss": 1.76337890625, "step": 1580 }, { "epoch": 1.264, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6806640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1580 }, { "epoch": 1.2648, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6806640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1581 }, { "epoch": 1.2656, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13720703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6806640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1582 }, { "epoch": 1.2664, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.378173828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.047119140625, "step": 1583 }, { "epoch": 1.2671999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046875, "step": 1584 }, { "epoch": 1.268, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.416015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046875, "step": 1585 }, { "epoch": 1.2688, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046875, "step": 1586 }, { "epoch": 1.2696, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046875, "step": 1587 }, { "epoch": 1.2704, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046875, "step": 1588 }, { "epoch": 1.2711999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.68017578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046630859375, "step": 1589 }, { "epoch": 1.272, "grad_norm": 0.16451396048069, "learning_rate": 2.88e-05, "loss": 1.76044921875, "step": 1590 }, { "epoch": 1.272, "model.model.language_model.layers.26.mlp.down_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.208984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.1708984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046630859375, "step": 1590 }, { "epoch": 1.2728, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046630859375, "step": 1591 }, { "epoch": 1.2736, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046630859375, "step": 1592 }, { "epoch": 1.2744, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046630859375, "step": 1593 }, { "epoch": 1.2752, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1594 }, { "epoch": 1.276, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.136474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1595 }, { "epoch": 1.2768, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1596 }, { "epoch": 1.2776, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1597 }, { "epoch": 1.2784, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1598 }, { "epoch": 1.2792, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1599 }, { "epoch": 1.28, "grad_norm": 0.15458697080612183, "learning_rate": 2.8666666666666668e-05, "loss": 1.77744140625, "step": 1600 }, { "epoch": 1.28, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.843505859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1600 }, { "epoch": 1.2808, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.279296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377685546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84326171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1601 }, { "epoch": 1.2816, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1602 }, { "epoch": 1.2824, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13623046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1603 }, { "epoch": 1.2832, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782958984375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41552734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1604 }, { "epoch": 1.284, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78271484375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1605 }, { "epoch": 1.2848, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78271484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1606 }, { "epoch": 1.2856, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78271484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1607 }, { "epoch": 1.2864, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37744140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1608 }, { "epoch": 1.2872, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135986328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1609 }, { "epoch": 1.288, "grad_norm": 0.14973245561122894, "learning_rate": 2.8533333333333333e-05, "loss": 1.78544921875, "step": 1610 }, { "epoch": 1.288, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1357421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1610 }, { "epoch": 1.2888, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1611 }, { "epoch": 1.2896, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.679931640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.377197265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1612 }, { "epoch": 1.2904, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1613 }, { "epoch": 1.2912, "model.model.language_model.layers.26.mlp.down_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.1650390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1614 }, { "epoch": 1.292, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1615 }, { "epoch": 1.2928, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8427734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1616 }, { "epoch": 1.2936, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1617 }, { "epoch": 1.2944, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135498046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1618 }, { "epoch": 1.2952, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1619 }, { "epoch": 1.296, "grad_norm": 0.1598503589630127, "learning_rate": 2.84e-05, "loss": 1.75703125, "step": 1620 }, { "epoch": 1.296, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1620 }, { "epoch": 1.2968, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.782470703125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376708984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1621 }, { "epoch": 1.2976, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376708984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1622 }, { "epoch": 1.2984, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1623 }, { "epoch": 1.2992, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7822265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1624 }, { "epoch": 1.3, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1625 }, { "epoch": 1.3008, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1626 }, { "epoch": 1.3016, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.415283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1627 }, { "epoch": 1.3024, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1628 }, { "epoch": 1.3032, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1629 }, { "epoch": 1.304, "grad_norm": 0.1786462813615799, "learning_rate": 2.8266666666666668e-05, "loss": 1.80546875, "step": 1630 }, { "epoch": 1.304, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1630 }, { "epoch": 1.3048, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13525390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1631 }, { "epoch": 1.3056, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1632 }, { "epoch": 1.3064, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1633 }, { "epoch": 1.3072, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1634 }, { "epoch": 1.308, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1635 }, { "epoch": 1.3088, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1636 }, { "epoch": 1.3096, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.135009765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1637 }, { "epoch": 1.3104, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1638 }, { "epoch": 1.3112, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4150390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1639 }, { "epoch": 1.312, "grad_norm": 0.19179297983646393, "learning_rate": 2.8133333333333334e-05, "loss": 1.759375, "step": 1640 }, { "epoch": 1.312, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1640 }, { "epoch": 1.3128, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1641 }, { "epoch": 1.3136, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1642 }, { "epoch": 1.3144, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67919921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37646484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1643 }, { "epoch": 1.3152, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1644 }, { "epoch": 1.316, "model.model.language_model.layers.26.mlp.down_proj": 0.38671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134521484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.345703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781982421875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1645 }, { "epoch": 1.3168, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1646 }, { "epoch": 1.3176, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1647 }, { "epoch": 1.3184, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1648 }, { "epoch": 1.3192, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.376220703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1649 }, { "epoch": 1.32, "grad_norm": 0.15814615786075592, "learning_rate": 2.8000000000000003e-05, "loss": 1.7779296875, "step": 1650 }, { "epoch": 1.32, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1650 }, { "epoch": 1.3208, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1651 }, { "epoch": 1.3216, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1652 }, { "epoch": 1.3224, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1653 }, { "epoch": 1.3232, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1654 }, { "epoch": 1.324, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414794921875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678955078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842529296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1655 }, { "epoch": 1.3248, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84228515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1656 }, { "epoch": 1.3256000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84228515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1657 }, { "epoch": 1.3264, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1658 }, { "epoch": 1.3272, "model.model.language_model.layers.26.mlp.down_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1659 }, { "epoch": 1.328, "grad_norm": 0.1802864372730255, "learning_rate": 2.786666666666667e-05, "loss": 1.73984375, "step": 1660 }, { "epoch": 1.328, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1660 }, { "epoch": 1.3288, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1661 }, { "epoch": 1.3296000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6787109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.842041015625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1662 }, { "epoch": 1.3304, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1663 }, { "epoch": 1.3312, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1664 }, { "epoch": 1.332, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375732421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1665 }, { "epoch": 1.3328, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1666 }, { "epoch": 1.3336000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1667 }, { "epoch": 1.3344, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.38671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1668 }, { "epoch": 1.3352, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1669 }, { "epoch": 1.336, "grad_norm": 0.17818385362625122, "learning_rate": 2.7733333333333334e-05, "loss": 1.73232421875, "step": 1670 }, { "epoch": 1.336, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1670 }, { "epoch": 1.3368, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1671 }, { "epoch": 1.3376000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1672 }, { "epoch": 1.3384, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1673 }, { "epoch": 1.3392, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1674 }, { "epoch": 1.34, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1675 }, { "epoch": 1.3408, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1676 }, { "epoch": 1.3416000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1677 }, { "epoch": 1.3424, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1678 }, { "epoch": 1.3432, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1679 }, { "epoch": 1.3439999999999999, "grad_norm": 0.15910208225250244, "learning_rate": 2.7600000000000003e-05, "loss": 1.7443359375, "step": 1680 }, { "epoch": 1.3439999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1680 }, { "epoch": 1.3448, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1681 }, { "epoch": 1.3456000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.678466796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1682 }, { "epoch": 1.3464, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13427734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41455078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1683 }, { "epoch": 1.3472, "model.model.language_model.layers.26.mlp.down_proj": 0.384765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1684 }, { "epoch": 1.3479999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1685 }, { "epoch": 1.3488, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1686 }, { "epoch": 1.3496000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1687 }, { "epoch": 1.3504, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1688 }, { "epoch": 1.3512, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1689 }, { "epoch": 1.3519999999999999, "grad_norm": 0.1812092512845993, "learning_rate": 2.746666666666667e-05, "loss": 1.73076171875, "step": 1690 }, { "epoch": 1.3519999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.134033203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1748046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1690 }, { "epoch": 1.3528, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1691 }, { "epoch": 1.3536000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67822265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375244140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1692 }, { "epoch": 1.3544, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.21875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1693 }, { "epoch": 1.3552, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1694 }, { "epoch": 1.3559999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1695 }, { "epoch": 1.3568, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1696 }, { "epoch": 1.3576, "model.model.language_model.layers.26.mlp.down_proj": 0.3671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1697 }, { "epoch": 1.3584, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1698 }, { "epoch": 1.3592, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1699 }, { "epoch": 1.3599999999999999, "grad_norm": 0.16335253417491913, "learning_rate": 2.733333333333333e-05, "loss": 1.75185546875, "step": 1700 }, { "epoch": 1.3599999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.133544921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1700 }, { "epoch": 1.3608, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.133544921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1701 }, { "epoch": 1.3616, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.133056640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374755859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1702 }, { "epoch": 1.3624, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.133056640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37451171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1703 }, { "epoch": 1.3632, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37451171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1704 }, { "epoch": 1.3639999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37451171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1705 }, { "epoch": 1.3648, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1706 }, { "epoch": 1.3656, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1707 }, { "epoch": 1.3664, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1708 }, { "epoch": 1.3672, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1709 }, { "epoch": 1.3679999999999999, "grad_norm": 0.15317372977733612, "learning_rate": 2.7200000000000004e-05, "loss": 1.74716796875, "step": 1710 }, { "epoch": 1.3679999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1710 }, { "epoch": 1.3688, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1711 }, { "epoch": 1.3696, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1712 }, { "epoch": 1.3704, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1713 }, { "epoch": 1.3712, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1714 }, { "epoch": 1.3719999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1715 }, { "epoch": 1.3728, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1716 }, { "epoch": 1.3736, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1717 }, { "epoch": 1.3744, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.414306640625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1718 }, { "epoch": 1.3752, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1719 }, { "epoch": 1.376, "grad_norm": 0.18508222699165344, "learning_rate": 2.706666666666667e-05, "loss": 1.7451171875, "step": 1720 }, { "epoch": 1.376, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1720 }, { "epoch": 1.3768, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1721 }, { "epoch": 1.3776, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1722 }, { "epoch": 1.3784, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1723 }, { "epoch": 1.3792, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132568359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.374267578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1724 }, { "epoch": 1.38, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1725 }, { "epoch": 1.3808, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1726 }, { "epoch": 1.3816, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1727 }, { "epoch": 1.3824, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1728 }, { "epoch": 1.3832, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1729 }, { "epoch": 1.384, "grad_norm": 0.15611444413661957, "learning_rate": 2.6933333333333332e-05, "loss": 1.7662109375, "step": 1730 }, { "epoch": 1.384, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1730 }, { "epoch": 1.3848, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1731 }, { "epoch": 1.3856, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1732 }, { "epoch": 1.3864, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13232421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1733 }, { "epoch": 1.3872, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1734 }, { "epoch": 1.388, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1735 }, { "epoch": 1.3888, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1736 }, { "epoch": 1.3896, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78173828125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1737 }, { "epoch": 1.3904, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3740234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1738 }, { "epoch": 1.3912, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373779296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1739 }, { "epoch": 1.392, "grad_norm": 0.1452898383140564, "learning_rate": 2.6800000000000004e-05, "loss": 1.73515625, "step": 1740 }, { "epoch": 1.392, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1740 }, { "epoch": 1.3928, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1741 }, { "epoch": 1.3936, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1742 }, { "epoch": 1.3944, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1743 }, { "epoch": 1.3952, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.132080078125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1744 }, { "epoch": 1.396, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677490234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1745 }, { "epoch": 1.3968, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1746 }, { "epoch": 1.3976, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04638671875, "step": 1747 }, { "epoch": 1.3984, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1748 }, { "epoch": 1.3992, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1749 }, { "epoch": 1.4, "grad_norm": 0.18338313698768616, "learning_rate": 2.6666666666666667e-05, "loss": 1.72666015625, "step": 1750 }, { "epoch": 1.4, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1750 }, { "epoch": 1.4008, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1751 }, { "epoch": 1.4016, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1752 }, { "epoch": 1.4024, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1753 }, { "epoch": 1.4032, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1754 }, { "epoch": 1.404, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781494140625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1755 }, { "epoch": 1.4048, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1756 }, { "epoch": 1.4056, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1757 }, { "epoch": 1.4064, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373291015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1758 }, { "epoch": 1.4072, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1759 }, { "epoch": 1.408, "grad_norm": 0.1422826498746872, "learning_rate": 2.6533333333333332e-05, "loss": 1.76201171875, "step": 1760 }, { "epoch": 1.408, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1760 }, { "epoch": 1.4088, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1761 }, { "epoch": 1.4096, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1762 }, { "epoch": 1.4104, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1763 }, { "epoch": 1.4112, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1764 }, { "epoch": 1.412, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1765 }, { "epoch": 1.4128, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1766 }, { "epoch": 1.4136, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1767 }, { "epoch": 1.4144, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1768 }, { "epoch": 1.4152, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841552734375, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1769 }, { "epoch": 1.416, "grad_norm": 0.1633981466293335, "learning_rate": 2.64e-05, "loss": 1.71728515625, "step": 1770 }, { "epoch": 1.416, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1770 }, { "epoch": 1.4168, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1771 }, { "epoch": 1.4176, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.677001953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1772 }, { "epoch": 1.4184, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1773 }, { "epoch": 1.4192, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.373046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1774 }, { "epoch": 1.42, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372802734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1775 }, { "epoch": 1.4208, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6767578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372802734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1776 }, { "epoch": 1.4216, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.22265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1777 }, { "epoch": 1.4224, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1778 }, { "epoch": 1.4232, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1779 }, { "epoch": 1.424, "grad_norm": 0.14626099169254303, "learning_rate": 2.6266666666666667e-05, "loss": 1.7048828125, "step": 1780 }, { "epoch": 1.424, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1318359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67626953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1780 }, { "epoch": 1.4248, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.676025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1781 }, { "epoch": 1.4256, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.676025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1782 }, { "epoch": 1.4264000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.676025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37255859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1783 }, { "epoch": 1.4272, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.676025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1784 }, { "epoch": 1.428, "model.model.language_model.layers.26.mlp.down_proj": 0.234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.676025390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1785 }, { "epoch": 1.4288, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131591796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1786 }, { "epoch": 1.4296, "model.model.language_model.layers.26.mlp.down_proj": 0.35546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1787 }, { "epoch": 1.4304000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1788 }, { "epoch": 1.4312, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1789 }, { "epoch": 1.432, "grad_norm": 0.14473386108875275, "learning_rate": 2.6133333333333333e-05, "loss": 1.7296875, "step": 1790 }, { "epoch": 1.432, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1790 }, { "epoch": 1.4328, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1791 }, { "epoch": 1.4336, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1792 }, { "epoch": 1.4344000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.372314453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1793 }, { "epoch": 1.4352, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1794 }, { "epoch": 1.436, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1795 }, { "epoch": 1.4368, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13134765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1796 }, { "epoch": 1.4376, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.216796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84130859375, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1797 }, { "epoch": 1.4384000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1798 }, { "epoch": 1.4392, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.34765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.33203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1799 }, { "epoch": 1.44, "grad_norm": 0.165522500872612, "learning_rate": 2.6000000000000002e-05, "loss": 1.7271484375, "step": 1800 }, { "epoch": 1.44, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1800 }, { "epoch": 1.4408, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1801 }, { "epoch": 1.4416, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1802 }, { "epoch": 1.4424000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1803 }, { "epoch": 1.4432, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.781005859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.046142578125, "step": 1804 }, { "epoch": 1.444, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1805 }, { "epoch": 1.4447999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1806 }, { "epoch": 1.4456, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1807 }, { "epoch": 1.4464000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1808 }, { "epoch": 1.4472, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1809 }, { "epoch": 1.448, "grad_norm": 0.14275521039962769, "learning_rate": 2.5866666666666667e-05, "loss": 1.7388671875, "step": 1810 }, { "epoch": 1.448, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3720703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1810 }, { "epoch": 1.4487999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1811 }, { "epoch": 1.4496, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1812 }, { "epoch": 1.4504000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1813 }, { "epoch": 1.4512, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1814 }, { "epoch": 1.452, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371826171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1815 }, { "epoch": 1.4527999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1816 }, { "epoch": 1.4536, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1817 }, { "epoch": 1.4544000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1818 }, { "epoch": 1.4552, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1819 }, { "epoch": 1.456, "grad_norm": 0.16061963140964508, "learning_rate": 2.5733333333333337e-05, "loss": 1.731640625, "step": 1820 }, { "epoch": 1.456, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1820 }, { "epoch": 1.4567999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.131103515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1821 }, { "epoch": 1.4576, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1822 }, { "epoch": 1.4584, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1823 }, { "epoch": 1.4592, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413818359375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37158203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1824 }, { "epoch": 1.46, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.371337890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1825 }, { "epoch": 1.4607999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.78076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1826 }, { "epoch": 1.4616, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.841064453125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1827 }, { "epoch": 1.4624, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1828 }, { "epoch": 1.4632, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1829 }, { "epoch": 1.464, "grad_norm": 0.14270153641700745, "learning_rate": 2.5600000000000002e-05, "loss": 1.71513671875, "step": 1830 }, { "epoch": 1.464, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1830 }, { "epoch": 1.4647999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1831 }, { "epoch": 1.4656, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8408203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1832 }, { "epoch": 1.4664, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1833 }, { "epoch": 1.4672, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675537109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1834 }, { "epoch": 1.468, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67529296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1835 }, { "epoch": 1.4687999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67529296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1836 }, { "epoch": 1.4696, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67529296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1837 }, { "epoch": 1.4704, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67529296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1838 }, { "epoch": 1.4712, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675048828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1839 }, { "epoch": 1.472, "grad_norm": 0.1680900603532791, "learning_rate": 2.5466666666666668e-05, "loss": 1.73642578125, "step": 1840 }, { "epoch": 1.472, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675048828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1840 }, { "epoch": 1.4727999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.675048828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1841 }, { "epoch": 1.4736, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1842 }, { "epoch": 1.4744, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1843 }, { "epoch": 1.4752, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1844 }, { "epoch": 1.476, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1845 }, { "epoch": 1.4768, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1846 }, { "epoch": 1.4776, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1847 }, { "epoch": 1.4784, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1848 }, { "epoch": 1.4792, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674560546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1849 }, { "epoch": 1.48, "grad_norm": 0.14432130753993988, "learning_rate": 2.5333333333333337e-05, "loss": 1.71552734375, "step": 1850 }, { "epoch": 1.48, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674560546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1850 }, { "epoch": 1.4808, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674560546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1851 }, { "epoch": 1.4816, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1852 }, { "epoch": 1.4824, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1853 }, { "epoch": 1.4832, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1854 }, { "epoch": 1.484, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1855 }, { "epoch": 1.4848, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370849609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1856 }, { "epoch": 1.4856, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1857 }, { "epoch": 1.4864, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130615234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1858 }, { "epoch": 1.4872, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1859 }, { "epoch": 1.488, "grad_norm": 0.1481044441461563, "learning_rate": 2.5200000000000003e-05, "loss": 1.73046875, "step": 1860 }, { "epoch": 1.488, "model.model.language_model.layers.26.mlp.down_proj": 0.216796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1860 }, { "epoch": 1.4888, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67431640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1861 }, { "epoch": 1.4896, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674072265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1862 }, { "epoch": 1.4904, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674072265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1863 }, { "epoch": 1.4912, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674072265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1864 }, { "epoch": 1.492, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.674072265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840576171875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1865 }, { "epoch": 1.4928, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.37060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1866 }, { "epoch": 1.4936, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1867 }, { "epoch": 1.4944, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1868 }, { "epoch": 1.4952, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1869 }, { "epoch": 1.496, "grad_norm": 0.1440674066543579, "learning_rate": 2.5066666666666665e-05, "loss": 1.676953125, "step": 1870 }, { "epoch": 1.496, "model.model.language_model.layers.26.mlp.down_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.228515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.17578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0458984375, "step": 1870 }, { "epoch": 1.4968, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1871 }, { "epoch": 1.4976, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1872 }, { "epoch": 1.4984, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1873 }, { "epoch": 1.4992, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1874 }, { "epoch": 1.5, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.370361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1875 }, { "epoch": 1.5008, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.13037109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.224609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1876 }, { "epoch": 1.5016, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1877 }, { "epoch": 1.5024, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1878 }, { "epoch": 1.5032, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780517578125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1879 }, { "epoch": 1.504, "grad_norm": 0.16179159283638, "learning_rate": 2.4933333333333334e-05, "loss": 1.70947265625, "step": 1880 }, { "epoch": 1.504, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1880 }, { "epoch": 1.5048, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1881 }, { "epoch": 1.5056, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.130126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1882 }, { "epoch": 1.5064, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1883 }, { "epoch": 1.5072, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1884 }, { "epoch": 1.508, "model.model.language_model.layers.26.mlp.down_proj": 0.236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1885 }, { "epoch": 1.5088, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1886 }, { "epoch": 1.5096, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1887 }, { "epoch": 1.5104, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1888 }, { "epoch": 1.5112, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7802734375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1889 }, { "epoch": 1.512, "grad_norm": 0.16700120270252228, "learning_rate": 2.48e-05, "loss": 1.707421875, "step": 1890 }, { "epoch": 1.512, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673583984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1890 }, { "epoch": 1.5128, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1891 }, { "epoch": 1.5135999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1892 }, { "epoch": 1.5144, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1893 }, { "epoch": 1.5152, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1894 }, { "epoch": 1.516, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1895 }, { "epoch": 1.5168, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.780029296875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1896 }, { "epoch": 1.5175999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77978515625, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1897 }, { "epoch": 1.5184, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77978515625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1898 }, { "epoch": 1.5192, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77978515625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1899 }, { "epoch": 1.52, "grad_norm": 0.1691533774137497, "learning_rate": 2.466666666666667e-05, "loss": 1.71220703125, "step": 1900 }, { "epoch": 1.52, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1298828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1900 }, { "epoch": 1.5208, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129638671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1901 }, { "epoch": 1.5215999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129638671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1902 }, { "epoch": 1.5224, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129638671875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.84033203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1903 }, { "epoch": 1.5232, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1904 }, { "epoch": 1.524, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41357421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1905 }, { "epoch": 1.5248, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1906 }, { "epoch": 1.5255999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1907 }, { "epoch": 1.5264, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12939453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1908 }, { "epoch": 1.5272000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1909 }, { "epoch": 1.528, "grad_norm": 0.14886461198329926, "learning_rate": 2.4533333333333334e-05, "loss": 1.71103515625, "step": 1910 }, { "epoch": 1.528, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1910 }, { "epoch": 1.5288, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3701171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1911 }, { "epoch": 1.5295999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1912 }, { "epoch": 1.5304, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1913 }, { "epoch": 1.5312000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1914 }, { "epoch": 1.532, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1915 }, { "epoch": 1.5328, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.337890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1916 }, { "epoch": 1.5335999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1917 }, { "epoch": 1.5344, "model.model.language_model.layers.26.mlp.down_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.413330078125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1918 }, { "epoch": 1.5352000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.840087890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1919 }, { "epoch": 1.536, "grad_norm": 0.1483919620513916, "learning_rate": 2.44e-05, "loss": 1.69970703125, "step": 1920 }, { "epoch": 1.536, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1920 }, { "epoch": 1.5368, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1921 }, { "epoch": 1.5375999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1922 }, { "epoch": 1.5384, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1923 }, { "epoch": 1.5392000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1924 }, { "epoch": 1.54, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1925 }, { "epoch": 1.5408, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1926 }, { "epoch": 1.5415999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.129150390625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1927 }, { "epoch": 1.5424, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67333984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369873046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839599609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1928 }, { "epoch": 1.5432000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673095703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36962890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045654296875, "step": 1929 }, { "epoch": 1.544, "grad_norm": 0.15542861819267273, "learning_rate": 2.426666666666667e-05, "loss": 1.7154296875, "step": 1930 }, { "epoch": 1.544, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.673095703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36962890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04541015625, "step": 1930 }, { "epoch": 1.5448, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369384765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04541015625, "step": 1931 }, { "epoch": 1.5455999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779541015625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369384765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04541015625, "step": 1932 }, { "epoch": 1.5464, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369384765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1933 }, { "epoch": 1.5472000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1934 }, { "epoch": 1.548, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.369140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1935 }, { "epoch": 1.5488, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368896484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1936 }, { "epoch": 1.5495999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368896484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1937 }, { "epoch": 1.5504, "model.model.language_model.layers.26.mlp.down_proj": 0.373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1938 }, { "epoch": 1.5512000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1939 }, { "epoch": 1.552, "grad_norm": 0.14198905229568481, "learning_rate": 2.4133333333333335e-05, "loss": 1.70380859375, "step": 1940 }, { "epoch": 1.552, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1940 }, { "epoch": 1.5528, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1941 }, { "epoch": 1.5535999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.045166015625, "step": 1942 }, { "epoch": 1.5544, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044921875, "step": 1943 }, { "epoch": 1.5552000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044921875, "step": 1944 }, { "epoch": 1.556, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044921875, "step": 1945 }, { "epoch": 1.5568, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83935546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1946 }, { "epoch": 1.5575999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1947 }, { "epoch": 1.5584, "model.model.language_model.layers.26.mlp.down_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.1826171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4130859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.12890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1948 }, { "epoch": 1.5592000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1949 }, { "epoch": 1.56, "grad_norm": 0.15828800201416016, "learning_rate": 2.4e-05, "loss": 1.73642578125, "step": 1950 }, { "epoch": 1.56, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128662109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1950 }, { "epoch": 1.5608, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128662109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1951 }, { "epoch": 1.5615999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36865234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1952 }, { "epoch": 1.5624, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12841796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1953 }, { "epoch": 1.5632000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779296875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.839111328125, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1954 }, { "epoch": 1.564, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1787109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6728515625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1955 }, { "epoch": 1.5648, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672607421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.368408203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1956 }, { "epoch": 1.5655999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672607421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1957 }, { "epoch": 1.5664, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672607421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1958 }, { "epoch": 1.5672000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1959 }, { "epoch": 1.568, "grad_norm": 0.16219857335090637, "learning_rate": 2.3866666666666666e-05, "loss": 1.71494140625, "step": 1960 }, { "epoch": 1.568, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412841796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1960 }, { "epoch": 1.5688, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1961 }, { "epoch": 1.5695999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41259765625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1962 }, { "epoch": 1.5704, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3681640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1963 }, { "epoch": 1.5712000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.128173828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1964 }, { "epoch": 1.572, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1965 }, { "epoch": 1.5728, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1966 }, { "epoch": 1.5735999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1967 }, { "epoch": 1.5744, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1968 }, { "epoch": 1.5752000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.672119140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1969 }, { "epoch": 1.576, "grad_norm": 0.16721494495868683, "learning_rate": 2.3733333333333335e-05, "loss": 1.6685546875, "step": 1970 }, { "epoch": 1.576, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1970 }, { "epoch": 1.5768, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1971 }, { "epoch": 1.5776, "model.model.language_model.layers.26.mlp.down_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367919921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1972 }, { "epoch": 1.5784, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1973 }, { "epoch": 1.5792000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1974 }, { "epoch": 1.58, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1975 }, { "epoch": 1.5808, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1279296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1976 }, { "epoch": 1.5816, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1977 }, { "epoch": 1.5824, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412353515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1978 }, { "epoch": 1.5832000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.22265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.181640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1979 }, { "epoch": 1.584, "grad_norm": 0.14237193763256073, "learning_rate": 2.36e-05, "loss": 1.71474609375, "step": 1980 }, { "epoch": 1.584, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1980 }, { "epoch": 1.5848, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36767578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1981 }, { "epoch": 1.5856, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.28125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1982 }, { "epoch": 1.5864, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1983 }, { "epoch": 1.5872000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12744140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1984 }, { "epoch": 1.588, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1985 }, { "epoch": 1.5888, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67138671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1986 }, { "epoch": 1.5896, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.671142578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.367431640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1987 }, { "epoch": 1.5904, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.412109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1988 }, { "epoch": 1.5912, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1989 }, { "epoch": 1.592, "grad_norm": 0.14856356382369995, "learning_rate": 2.3466666666666667e-05, "loss": 1.6984375, "step": 1990 }, { "epoch": 1.592, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1990 }, { "epoch": 1.5928, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1991 }, { "epoch": 1.5936, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411865234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1992 }, { "epoch": 1.5944, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1993 }, { "epoch": 1.5952, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6708984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1994 }, { "epoch": 1.596, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1995 }, { "epoch": 1.5968, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8388671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1996 }, { "epoch": 1.5976, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.127197265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1997 }, { "epoch": 1.5984, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1998 }, { "epoch": 1.5992, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 1999 }, { "epoch": 1.6, "grad_norm": 0.15947073698043823, "learning_rate": 2.3333333333333336e-05, "loss": 1.6771484375, "step": 2000 }, { "epoch": 1.6, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2000 }, { "epoch": 1.6008, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2001 }, { "epoch": 1.6016, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2002 }, { "epoch": 1.6024, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2003 }, { "epoch": 1.6032, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2004 }, { "epoch": 1.604, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2005 }, { "epoch": 1.6048, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044677734375, "step": 2006 }, { "epoch": 1.6056, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2007 }, { "epoch": 1.6064, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2008 }, { "epoch": 1.6072, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36669921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2009 }, { "epoch": 1.608, "grad_norm": 0.14622269570827484, "learning_rate": 2.32e-05, "loss": 1.7232421875, "step": 2010 }, { "epoch": 1.608, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.366455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838623046875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2010 }, { "epoch": 1.6088, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.366455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2011 }, { "epoch": 1.6096, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.366455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2012 }, { "epoch": 1.6104, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.366455078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2013 }, { "epoch": 1.6112, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04443359375, "step": 2014 }, { "epoch": 1.612, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2015 }, { "epoch": 1.6128, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.779052734375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2016 }, { "epoch": 1.6136, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2017 }, { "epoch": 1.6143999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2018 }, { "epoch": 1.6152, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41162109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2019 }, { "epoch": 1.616, "grad_norm": 0.14740265905857086, "learning_rate": 2.3066666666666667e-05, "loss": 1.70107421875, "step": 2020 }, { "epoch": 1.616, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2020 }, { "epoch": 1.6168, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.330078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2021 }, { "epoch": 1.6176, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126708984375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2022 }, { "epoch": 1.6183999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2023 }, { "epoch": 1.6192, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2024 }, { "epoch": 1.62, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77880859375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.411376953125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2025 }, { "epoch": 1.6208, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2026 }, { "epoch": 1.6216, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2027 }, { "epoch": 1.6223999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.838134765625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2028 }, { "epoch": 1.6232, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2029 }, { "epoch": 1.624, "grad_norm": 0.1375412791967392, "learning_rate": 2.2933333333333333e-05, "loss": 1.71513671875, "step": 2030 }, { "epoch": 1.624, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2030 }, { "epoch": 1.6248, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2031 }, { "epoch": 1.6256, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2032 }, { "epoch": 1.6263999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12646484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2033 }, { "epoch": 1.6272, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.126220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.044189453125, "step": 2034 }, { "epoch": 1.6280000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3662109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2035 }, { "epoch": 1.6288, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365966796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2036 }, { "epoch": 1.6296, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365966796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2037 }, { "epoch": 1.6303999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2038 }, { "epoch": 1.6312, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2039 }, { "epoch": 1.6320000000000001, "grad_norm": 0.15841689705848694, "learning_rate": 2.2800000000000002e-05, "loss": 1.6974609375, "step": 2040 }, { "epoch": 1.6320000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2040 }, { "epoch": 1.6328, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2041 }, { "epoch": 1.6336, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0439453125, "step": 2042 }, { "epoch": 1.6343999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2043 }, { "epoch": 1.6352, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2044 }, { "epoch": 1.6360000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2045 }, { "epoch": 1.6368, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1259765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.29296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.40234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2046 }, { "epoch": 1.6376, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125732421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2047 }, { "epoch": 1.6383999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125732421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2048 }, { "epoch": 1.6392, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125732421875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2049 }, { "epoch": 1.6400000000000001, "grad_norm": 0.15715304017066956, "learning_rate": 2.2666666666666668e-05, "loss": 1.6662109375, "step": 2050 }, { "epoch": 1.6400000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837890625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2050 }, { "epoch": 1.6408, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2051 }, { "epoch": 1.6416, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12548828125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2052 }, { "epoch": 1.6423999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2053 }, { "epoch": 1.6432, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365478515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2054 }, { "epoch": 1.6440000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2055 }, { "epoch": 1.6448, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2056 }, { "epoch": 1.6456, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2057 }, { "epoch": 1.6463999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670654296875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2058 }, { "epoch": 1.6472, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2059 }, { "epoch": 1.6480000000000001, "grad_norm": 0.15639474987983704, "learning_rate": 2.2533333333333333e-05, "loss": 1.68115234375, "step": 2060 }, { "epoch": 1.6480000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2060 }, { "epoch": 1.6488, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2061 }, { "epoch": 1.6496, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2062 }, { "epoch": 1.6503999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2063 }, { "epoch": 1.6512, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2064 }, { "epoch": 1.6520000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4111328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837646484375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2065 }, { "epoch": 1.6528, "model.model.language_model.layers.26.mlp.down_proj": 0.380859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2066 }, { "epoch": 1.6536, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2067 }, { "epoch": 1.6543999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2068 }, { "epoch": 1.6552, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2069 }, { "epoch": 1.6560000000000001, "grad_norm": 0.1622970551252365, "learning_rate": 2.2400000000000002e-05, "loss": 1.68291015625, "step": 2070 }, { "epoch": 1.6560000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2070 }, { "epoch": 1.6568, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2071 }, { "epoch": 1.6576, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2072 }, { "epoch": 1.6583999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2073 }, { "epoch": 1.6592, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2074 }, { "epoch": 1.6600000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2075 }, { "epoch": 1.6608, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2076 }, { "epoch": 1.6616, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2077 }, { "epoch": 1.6623999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2078 }, { "epoch": 1.6632, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2079 }, { "epoch": 1.6640000000000001, "grad_norm": 0.15390853583812714, "learning_rate": 2.2266666666666668e-05, "loss": 1.69365234375, "step": 2080 }, { "epoch": 1.6640000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2080 }, { "epoch": 1.6648, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2081 }, { "epoch": 1.6656, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2082 }, { "epoch": 1.6663999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364990234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2083 }, { "epoch": 1.6672, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2084 }, { "epoch": 1.6680000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410888671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2085 }, { "epoch": 1.6688, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125244140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.67041015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2086 }, { "epoch": 1.6696, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2087 }, { "epoch": 1.6703999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36474609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2088 }, { "epoch": 1.6712, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.670166015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2089 }, { "epoch": 1.6720000000000002, "grad_norm": 0.16721579432487488, "learning_rate": 2.2133333333333334e-05, "loss": 1.67490234375, "step": 2090 }, { "epoch": 1.6720000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2090 }, { "epoch": 1.6728, "model.model.language_model.layers.26.mlp.down_proj": 0.349609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2091 }, { "epoch": 1.6736, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778564453125, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364501953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2092 }, { "epoch": 1.6743999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2093 }, { "epoch": 1.6752, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.837158203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2094 }, { "epoch": 1.6760000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124755859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.364013671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2095 }, { "epoch": 1.6768, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.124267578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2096 }, { "epoch": 1.6776, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36376953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043701171875, "step": 2097 }, { "epoch": 1.6784, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.363525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2098 }, { "epoch": 1.6792, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.363525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2099 }, { "epoch": 1.6800000000000002, "grad_norm": 0.12993182241916656, "learning_rate": 2.2000000000000003e-05, "loss": 1.69169921875, "step": 2100 }, { "epoch": 1.6800000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.363525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2100 }, { "epoch": 1.6808, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.363525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8369140625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2101 }, { "epoch": 1.6816, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2102 }, { "epoch": 1.6824, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83642578125, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2103 }, { "epoch": 1.6832, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2104 }, { "epoch": 1.6840000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2105 }, { "epoch": 1.6848, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2106 }, { "epoch": 1.6856, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.778076171875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2107 }, { "epoch": 1.6864, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2108 }, { "epoch": 1.6872, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2109 }, { "epoch": 1.688, "grad_norm": 0.15012043714523315, "learning_rate": 2.186666666666667e-05, "loss": 1.6798828125, "step": 2110 }, { "epoch": 1.688, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.363037109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.836181640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2110 }, { "epoch": 1.6888, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.28125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36279296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2111 }, { "epoch": 1.6896, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123779296875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36279296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2112 }, { "epoch": 1.6904, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12353515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2113 }, { "epoch": 1.6912, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12353515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.41064453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2114 }, { "epoch": 1.692, "model.model.language_model.layers.26.mlp.down_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123291015625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2115 }, { "epoch": 1.6928, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2116 }, { "epoch": 1.6936, "model.model.language_model.layers.26.mlp.down_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669677734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362548828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2117 }, { "epoch": 1.6944, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2118 }, { "epoch": 1.6952, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2119 }, { "epoch": 1.696, "grad_norm": 0.15288235247135162, "learning_rate": 2.1733333333333334e-05, "loss": 1.66337890625, "step": 2120 }, { "epoch": 1.696, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2120 }, { "epoch": 1.6968, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2121 }, { "epoch": 1.6976, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.410400390625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2122 }, { "epoch": 1.6984, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2123 }, { "epoch": 1.6992, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2124 }, { "epoch": 1.7, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2125 }, { "epoch": 1.7008, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2126 }, { "epoch": 1.7016, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2127 }, { "epoch": 1.7024, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2128 }, { "epoch": 1.7032, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2129 }, { "epoch": 1.704, "grad_norm": 0.1366635411977768, "learning_rate": 2.16e-05, "loss": 1.65927734375, "step": 2130 }, { "epoch": 1.704, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2130 }, { "epoch": 1.7048, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2131 }, { "epoch": 1.7056, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2132 }, { "epoch": 1.7064, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2133 }, { "epoch": 1.7072, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2134 }, { "epoch": 1.708, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3623046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2135 }, { "epoch": 1.7088, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2136 }, { "epoch": 1.7096, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2137 }, { "epoch": 1.7104, "model.model.language_model.layers.26.mlp.down_proj": 0.203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77783203125, "model.model.language_model.layers.26.mlp.up_proj": 0.177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2138 }, { "epoch": 1.7112, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2139 }, { "epoch": 1.712, "grad_norm": 0.16390572488307953, "learning_rate": 2.146666666666667e-05, "loss": 1.65283203125, "step": 2140 }, { "epoch": 1.712, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2140 }, { "epoch": 1.7128, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2141 }, { "epoch": 1.7136, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2142 }, { "epoch": 1.7144, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2143 }, { "epoch": 1.7151999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2144 }, { "epoch": 1.716, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835693359375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2145 }, { "epoch": 1.7168, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66943359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2146 }, { "epoch": 1.7176, "model.model.language_model.layers.26.mlp.down_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669189453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2147 }, { "epoch": 1.7184, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669189453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2148 }, { "epoch": 1.7191999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.669189453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2149 }, { "epoch": 1.72, "grad_norm": 0.1391422152519226, "learning_rate": 2.1333333333333335e-05, "loss": 1.67734375, "step": 2150 }, { "epoch": 1.72, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6689453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2150 }, { "epoch": 1.7208, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2151 }, { "epoch": 1.7216, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2152 }, { "epoch": 1.7224, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2153 }, { "epoch": 1.7231999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409912109375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2154 }, { "epoch": 1.724, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.17578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2155 }, { "epoch": 1.7248, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.30078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.34375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2156 }, { "epoch": 1.7256, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.362060546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2157 }, { "epoch": 1.7264, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2158 }, { "epoch": 1.7271999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2159 }, { "epoch": 1.728, "grad_norm": 0.15731486678123474, "learning_rate": 2.12e-05, "loss": 1.6794921875, "step": 2160 }, { "epoch": 1.728, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40966796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2160 }, { "epoch": 1.7288000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2161 }, { "epoch": 1.7296, "model.model.language_model.layers.26.mlp.down_proj": 0.361328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77734375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2162 }, { "epoch": 1.7304, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2163 }, { "epoch": 1.7311999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.409423828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2164 }, { "epoch": 1.732, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66845703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2165 }, { "epoch": 1.7328000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.1826171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2166 }, { "epoch": 1.7336, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2167 }, { "epoch": 1.7344, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2168 }, { "epoch": 1.7351999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2169 }, { "epoch": 1.736, "grad_norm": 0.1488647609949112, "learning_rate": 2.106666666666667e-05, "loss": 1.675390625, "step": 2170 }, { "epoch": 1.736, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2170 }, { "epoch": 1.7368000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2171 }, { "epoch": 1.7376, "model.model.language_model.layers.26.mlp.down_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2172 }, { "epoch": 1.7384, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4091796875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2173 }, { "epoch": 1.7391999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.777099609375, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2174 }, { "epoch": 1.74, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36181640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2175 }, { "epoch": 1.7408000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2176 }, { "epoch": 1.7416, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2177 }, { "epoch": 1.7424, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.123046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2178 }, { "epoch": 1.7431999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2179 }, { "epoch": 1.744, "grad_norm": 0.16092383861541748, "learning_rate": 2.0933333333333335e-05, "loss": 1.6666015625, "step": 2180 }, { "epoch": 1.744, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2180 }, { "epoch": 1.7448000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408935546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361572265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2181 }, { "epoch": 1.7456, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40869140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2182 }, { "epoch": 1.7464, "model.model.language_model.layers.26.mlp.down_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40869140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2183 }, { "epoch": 1.7471999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40869140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2184 }, { "epoch": 1.748, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40869140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2185 }, { "epoch": 1.7488000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.232421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2186 }, { "epoch": 1.7496, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2187 }, { "epoch": 1.7504, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2188 }, { "epoch": 1.7511999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.668212890625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2189 }, { "epoch": 1.752, "grad_norm": 0.15649822354316711, "learning_rate": 2.08e-05, "loss": 1.6484375, "step": 2190 }, { "epoch": 1.752, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2190 }, { "epoch": 1.7528000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2191 }, { "epoch": 1.7536, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2192 }, { "epoch": 1.7544, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.22265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2193 }, { "epoch": 1.7551999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2194 }, { "epoch": 1.756, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2195 }, { "epoch": 1.7568000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2196 }, { "epoch": 1.7576, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2197 }, { "epoch": 1.7584, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83544921875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2198 }, { "epoch": 1.7591999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122802734375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2199 }, { "epoch": 1.76, "grad_norm": 0.1551268994808197, "learning_rate": 2.0666666666666666e-05, "loss": 1.6728515625, "step": 2200 }, { "epoch": 1.76, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04345703125, "step": 2200 }, { "epoch": 1.7608000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2201 }, { "epoch": 1.7616, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2202 }, { "epoch": 1.7624, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2203 }, { "epoch": 1.7631999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77685546875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2204 }, { "epoch": 1.764, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2205 }, { "epoch": 1.7648000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.043212890625, "step": 2206 }, { "epoch": 1.7656, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2207 }, { "epoch": 1.7664, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2208 }, { "epoch": 1.7671999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2209 }, { "epoch": 1.768, "grad_norm": 0.13985460996627808, "learning_rate": 2.0533333333333336e-05, "loss": 1.6568359375, "step": 2210 }, { "epoch": 1.768, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2210 }, { "epoch": 1.7688000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408447265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2211 }, { "epoch": 1.7696, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2212 }, { "epoch": 1.7704, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667724609375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2213 }, { "epoch": 1.7711999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2214 }, { "epoch": 1.772, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2215 }, { "epoch": 1.7728000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2216 }, { "epoch": 1.7736, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12255859375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2217 }, { "epoch": 1.7744, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2218 }, { "epoch": 1.7752, "model.model.language_model.layers.26.mlp.down_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2219 }, { "epoch": 1.776, "grad_norm": 0.14984413981437683, "learning_rate": 2.04e-05, "loss": 1.66025390625, "step": 2220 }, { "epoch": 1.776, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2220 }, { "epoch": 1.7768000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2221 }, { "epoch": 1.7776, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2222 }, { "epoch": 1.7784, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.361083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2223 }, { "epoch": 1.7792, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2224 }, { "epoch": 1.78, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.36083984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2225 }, { "epoch": 1.7808000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2226 }, { "epoch": 1.7816, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2227 }, { "epoch": 1.7824, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2228 }, { "epoch": 1.7832, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.337890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.287109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.35546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2229 }, { "epoch": 1.784, "grad_norm": 0.14541102945804596, "learning_rate": 2.0266666666666667e-05, "loss": 1.66064453125, "step": 2230 }, { "epoch": 1.784, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2230 }, { "epoch": 1.7848000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.122314453125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.667236328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.360595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2231 }, { "epoch": 1.7856, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2232 }, { "epoch": 1.7864, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2233 }, { "epoch": 1.7872, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2234 }, { "epoch": 1.788, "model.model.language_model.layers.26.mlp.down_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.408203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2235 }, { "epoch": 1.7888, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2236 }, { "epoch": 1.7896, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2237 }, { "epoch": 1.7904, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2238 }, { "epoch": 1.7912, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1220703125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2239 }, { "epoch": 1.792, "grad_norm": 0.15273037552833557, "learning_rate": 2.0133333333333336e-05, "loss": 1.640234375, "step": 2240 }, { "epoch": 1.792, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121826171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.835205078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2240 }, { "epoch": 1.7928, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121826171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2241 }, { "epoch": 1.7936, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2242 }, { "epoch": 1.7944, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2243 }, { "epoch": 1.7952, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6669921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2244 }, { "epoch": 1.796, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2245 }, { "epoch": 1.7968, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776611328125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2246 }, { "epoch": 1.7976, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35986328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2247 }, { "epoch": 1.7984, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2248 }, { "epoch": 1.7992, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2249 }, { "epoch": 1.8, "grad_norm": 0.1438436657190323, "learning_rate": 2e-05, "loss": 1.64853515625, "step": 2250 }, { "epoch": 1.8, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2250 }, { "epoch": 1.8008, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2251 }, { "epoch": 1.8016, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1884765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2252 }, { "epoch": 1.8024, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666748046875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2253 }, { "epoch": 1.8032, "model.model.language_model.layers.26.mlp.down_proj": 0.23046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407958984375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2254 }, { "epoch": 1.804, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2255 }, { "epoch": 1.8048, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2256 }, { "epoch": 1.8056, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2257 }, { "epoch": 1.8064, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2258 }, { "epoch": 1.8072, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2259 }, { "epoch": 1.808, "grad_norm": 0.16514648497104645, "learning_rate": 1.9866666666666667e-05, "loss": 1.69150390625, "step": 2260 }, { "epoch": 1.808, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2260 }, { "epoch": 1.8088, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2261 }, { "epoch": 1.8096, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2262 }, { "epoch": 1.8104, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2263 }, { "epoch": 1.8112, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2264 }, { "epoch": 1.812, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2265 }, { "epoch": 1.8128, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2266 }, { "epoch": 1.8136, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40771484375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2267 }, { "epoch": 1.8144, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2268 }, { "epoch": 1.8152, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2269 }, { "epoch": 1.8159999999999998, "grad_norm": 0.16208283603191376, "learning_rate": 1.9733333333333333e-05, "loss": 1.65029296875, "step": 2270 }, { "epoch": 1.8159999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.23046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2270 }, { "epoch": 1.8168, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2271 }, { "epoch": 1.8176, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2272 }, { "epoch": 1.8184, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2273 }, { "epoch": 1.8192, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2274 }, { "epoch": 1.8199999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2275 }, { "epoch": 1.8208, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2276 }, { "epoch": 1.8216, "model.model.language_model.layers.26.mlp.down_proj": 0.236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2277 }, { "epoch": 1.8224, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35888671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2278 }, { "epoch": 1.8232, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2279 }, { "epoch": 1.8239999999999998, "grad_norm": 0.15677478909492493, "learning_rate": 1.9600000000000002e-05, "loss": 1.623046875, "step": 2280 }, { "epoch": 1.8239999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04296875, "step": 2280 }, { "epoch": 1.8248, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12158203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2281 }, { "epoch": 1.8256000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2282 }, { "epoch": 1.8264, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2283 }, { "epoch": 1.8272, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.407470703125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2284 }, { "epoch": 1.8279999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2285 }, { "epoch": 1.8288, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.121337890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2286 }, { "epoch": 1.8296000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2287 }, { "epoch": 1.8304, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2288 }, { "epoch": 1.8312, "model.model.language_model.layers.26.mlp.down_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4072265625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2289 }, { "epoch": 1.8319999999999999, "grad_norm": 0.13851295411586761, "learning_rate": 1.9466666666666668e-05, "loss": 1.70087890625, "step": 2290 }, { "epoch": 1.8319999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120849609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2290 }, { "epoch": 1.8328, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120849609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2291 }, { "epoch": 1.8336000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2292 }, { "epoch": 1.8344, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2293 }, { "epoch": 1.8352, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2294 }, { "epoch": 1.8359999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2295 }, { "epoch": 1.8368, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2296 }, { "epoch": 1.8376000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2297 }, { "epoch": 1.8384, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358642578125, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2298 }, { "epoch": 1.8392, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2299 }, { "epoch": 1.8399999999999999, "grad_norm": 0.14050106704235077, "learning_rate": 1.9333333333333333e-05, "loss": 1.659375, "step": 2300 }, { "epoch": 1.8399999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2300 }, { "epoch": 1.8408, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2301 }, { "epoch": 1.8416000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2302 }, { "epoch": 1.8424, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2303 }, { "epoch": 1.8432, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2304 }, { "epoch": 1.8439999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3583984375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2305 }, { "epoch": 1.8448, "model.model.language_model.layers.26.mlp.down_proj": 0.3828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.349609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.37890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.28125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042724609375, "step": 2306 }, { "epoch": 1.8456000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2307 }, { "epoch": 1.8464, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406982421875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2308 }, { "epoch": 1.8472, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.12060546875, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.834716796875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2309 }, { "epoch": 1.8479999999999999, "grad_norm": 0.15339772403240204, "learning_rate": 1.9200000000000003e-05, "loss": 1.66337890625, "step": 2310 }, { "epoch": 1.8479999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.173828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2310 }, { "epoch": 1.8488, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.666015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2311 }, { "epoch": 1.8496000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2312 }, { "epoch": 1.8504, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2313 }, { "epoch": 1.8512, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2314 }, { "epoch": 1.8519999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2315 }, { "epoch": 1.8528, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2316 }, { "epoch": 1.8536000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.358154296875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2317 }, { "epoch": 1.8544, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.120361328125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2318 }, { "epoch": 1.8552, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2319 }, { "epoch": 1.8559999999999999, "grad_norm": 0.14502157270908356, "learning_rate": 1.9066666666666668e-05, "loss": 1.6462890625, "step": 2320 }, { "epoch": 1.8559999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2320 }, { "epoch": 1.8568, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2321 }, { "epoch": 1.8576000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2322 }, { "epoch": 1.8584, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2323 }, { "epoch": 1.8592, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2324 }, { "epoch": 1.8599999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2325 }, { "epoch": 1.8608, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2326 }, { "epoch": 1.8616000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2327 }, { "epoch": 1.8624, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2328 }, { "epoch": 1.8632, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2329 }, { "epoch": 1.8639999999999999, "grad_norm": 0.1417674869298935, "learning_rate": 1.8933333333333334e-05, "loss": 1.65888671875, "step": 2330 }, { "epoch": 1.8639999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.1787109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35791015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2330 }, { "epoch": 1.8648, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2331 }, { "epoch": 1.8656000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2332 }, { "epoch": 1.8664, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2333 }, { "epoch": 1.8672, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7763671875, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.665771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2334 }, { "epoch": 1.8679999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119873046875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2335 }, { "epoch": 1.8688, "model.model.language_model.layers.26.mlp.down_proj": 0.36328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2336 }, { "epoch": 1.8696000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2337 }, { "epoch": 1.8704, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2338 }, { "epoch": 1.8712, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2339 }, { "epoch": 1.8719999999999999, "grad_norm": 0.15082943439483643, "learning_rate": 1.88e-05, "loss": 1.65859375, "step": 2340 }, { "epoch": 1.8719999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2340 }, { "epoch": 1.8728, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2341 }, { "epoch": 1.8736000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.776123046875, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2342 }, { "epoch": 1.8744, "model.model.language_model.layers.26.mlp.down_proj": 0.234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2343 }, { "epoch": 1.8752, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2344 }, { "epoch": 1.876, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2345 }, { "epoch": 1.8768, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11962890625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2346 }, { "epoch": 1.8776000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119384765625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2347 }, { "epoch": 1.8784, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66552734375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2348 }, { "epoch": 1.8792, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2349 }, { "epoch": 1.88, "grad_norm": 0.14352259039878845, "learning_rate": 1.866666666666667e-05, "loss": 1.65810546875, "step": 2350 }, { "epoch": 1.88, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2350 }, { "epoch": 1.8808, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2351 }, { "epoch": 1.8816000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.228515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2352 }, { "epoch": 1.8824, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357666015625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2353 }, { "epoch": 1.8832, "model.model.language_model.layers.26.mlp.down_proj": 0.37890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.35546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2354 }, { "epoch": 1.884, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2355 }, { "epoch": 1.8848, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2356 }, { "epoch": 1.8856000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2357 }, { "epoch": 1.8864, "model.model.language_model.layers.26.mlp.down_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.23046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.17578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2358 }, { "epoch": 1.8872, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6650390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2359 }, { "epoch": 1.888, "grad_norm": 0.16105972230434418, "learning_rate": 1.8533333333333334e-05, "loss": 1.66552734375, "step": 2360 }, { "epoch": 1.888, "model.model.language_model.layers.26.mlp.down_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2360 }, { "epoch": 1.8888, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2361 }, { "epoch": 1.8896, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2362 }, { "epoch": 1.8904, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2363 }, { "epoch": 1.8912, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2364 }, { "epoch": 1.892, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2365 }, { "epoch": 1.8928, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2366 }, { "epoch": 1.8936, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2367 }, { "epoch": 1.8944, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2368 }, { "epoch": 1.8952, "model.model.language_model.layers.26.mlp.down_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1845703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2369 }, { "epoch": 1.896, "grad_norm": 0.15195605158805847, "learning_rate": 1.84e-05, "loss": 1.66865234375, "step": 2370 }, { "epoch": 1.896, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2370 }, { "epoch": 1.8968, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2371 }, { "epoch": 1.8976, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.119140625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2372 }, { "epoch": 1.8984, "model.model.language_model.layers.26.mlp.down_proj": 0.392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.369140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.302734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.341796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2373 }, { "epoch": 1.8992, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2374 }, { "epoch": 1.9, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2375 }, { "epoch": 1.9008, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2376 }, { "epoch": 1.9016, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2377 }, { "epoch": 1.9024, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2378 }, { "epoch": 1.9032, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2379 }, { "epoch": 1.904, "grad_norm": 0.15808890759944916, "learning_rate": 1.826666666666667e-05, "loss": 1.67421875, "step": 2380 }, { "epoch": 1.904, "model.model.language_model.layers.26.mlp.down_proj": 0.234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.357177734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2380 }, { "epoch": 1.9048, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35693359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2381 }, { "epoch": 1.9056, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2382 }, { "epoch": 1.9064, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2383 }, { "epoch": 1.9072, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2384 }, { "epoch": 1.908, "model.model.language_model.layers.26.mlp.down_proj": 0.244140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1884765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83447265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2385 }, { "epoch": 1.9088, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.326171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.834228515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2386 }, { "epoch": 1.9096, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.834228515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2387 }, { "epoch": 1.9104, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.834228515625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2388 }, { "epoch": 1.9112, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2389 }, { "epoch": 1.912, "grad_norm": 0.1408911496400833, "learning_rate": 1.8133333333333335e-05, "loss": 1.6537109375, "step": 2390 }, { "epoch": 1.912, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40673828125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2390 }, { "epoch": 1.9127999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2391 }, { "epoch": 1.9136, "model.model.language_model.layers.26.mlp.down_proj": 0.31640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2392 }, { "epoch": 1.9144, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2393 }, { "epoch": 1.9152, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2394 }, { "epoch": 1.916, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.228515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2395 }, { "epoch": 1.9167999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2396 }, { "epoch": 1.9176, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2397 }, { "epoch": 1.9184, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2398 }, { "epoch": 1.9192, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2399 }, { "epoch": 1.92, "grad_norm": 0.14048168063163757, "learning_rate": 1.8e-05, "loss": 1.6625, "step": 2400 }, { "epoch": 1.92, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2400 }, { "epoch": 1.9207999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118896484375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2401 }, { "epoch": 1.9216, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2402 }, { "epoch": 1.9224, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2403 }, { "epoch": 1.9232, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356689453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2404 }, { "epoch": 1.924, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2405 }, { "epoch": 1.9247999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2406 }, { "epoch": 1.9256, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2407 }, { "epoch": 1.9264000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2408 }, { "epoch": 1.9272, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2409 }, { "epoch": 1.928, "grad_norm": 0.13693244755268097, "learning_rate": 1.7866666666666666e-05, "loss": 1.6498046875, "step": 2410 }, { "epoch": 1.928, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406494140625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04248046875, "step": 2410 }, { "epoch": 1.9287999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2411 }, { "epoch": 1.9296, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77587890625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2412 }, { "epoch": 1.9304000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2413 }, { "epoch": 1.9312, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2414 }, { "epoch": 1.932, "model.model.language_model.layers.26.mlp.down_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2415 }, { "epoch": 1.9327999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2416 }, { "epoch": 1.9336, "model.model.language_model.layers.26.mlp.down_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2417 }, { "epoch": 1.9344000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2418 }, { "epoch": 1.9352, "model.model.language_model.layers.26.mlp.down_proj": 0.228515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1826171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2419 }, { "epoch": 1.936, "grad_norm": 0.15769335627555847, "learning_rate": 1.7733333333333335e-05, "loss": 1.6478515625, "step": 2420 }, { "epoch": 1.936, "model.model.language_model.layers.26.mlp.down_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2420 }, { "epoch": 1.9367999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.326171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2421 }, { "epoch": 1.9376, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2422 }, { "epoch": 1.9384000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11865234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2423 }, { "epoch": 1.9392, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2424 }, { "epoch": 1.94, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2425 }, { "epoch": 1.9407999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.32421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2426 }, { "epoch": 1.9416, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3564453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2427 }, { "epoch": 1.9424000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356201171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2428 }, { "epoch": 1.9432, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356201171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2429 }, { "epoch": 1.944, "grad_norm": 0.1852206587791443, "learning_rate": 1.76e-05, "loss": 1.650390625, "step": 2430 }, { "epoch": 1.944, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664794921875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.356201171875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2430 }, { "epoch": 1.9447999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.406005859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2431 }, { "epoch": 1.9456, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2432 }, { "epoch": 1.9464000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2433 }, { "epoch": 1.9472, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2434 }, { "epoch": 1.948, "model.model.language_model.layers.26.mlp.down_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1787109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35595703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2435 }, { "epoch": 1.9487999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2436 }, { "epoch": 1.9496, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.118408203125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2437 }, { "epoch": 1.9504000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2438 }, { "epoch": 1.9512, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833984375, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.042236328125, "step": 2439 }, { "epoch": 1.952, "grad_norm": 0.16946546733379364, "learning_rate": 1.7466666666666667e-05, "loss": 1.6365234375, "step": 2440 }, { "epoch": 1.952, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2440 }, { "epoch": 1.9527999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2441 }, { "epoch": 1.9536, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2442 }, { "epoch": 1.9544000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833740234375, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2443 }, { "epoch": 1.9552, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66455078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2444 }, { "epoch": 1.956, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2445 }, { "epoch": 1.9567999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2446 }, { "epoch": 1.9576, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2447 }, { "epoch": 1.9584000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2448 }, { "epoch": 1.9592, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2449 }, { "epoch": 1.96, "grad_norm": 0.13502874970436096, "learning_rate": 1.7333333333333336e-05, "loss": 1.62421875, "step": 2450 }, { "epoch": 1.96, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355712890625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2450 }, { "epoch": 1.9607999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40576171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2451 }, { "epoch": 1.9616, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2452 }, { "epoch": 1.9624000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2453 }, { "epoch": 1.9632, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2454 }, { "epoch": 1.964, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2455 }, { "epoch": 1.9647999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2456 }, { "epoch": 1.9656, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2457 }, { "epoch": 1.9664000000000001, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83349609375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0419921875, "step": 2458 }, { "epoch": 1.9672, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2459 }, { "epoch": 1.968, "grad_norm": 0.15218009054660797, "learning_rate": 1.7199999999999998e-05, "loss": 1.63642578125, "step": 2460 }, { "epoch": 1.968, "model.model.language_model.layers.26.mlp.down_proj": 0.236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1181640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.232421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1826171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2460 }, { "epoch": 1.9687999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.220703125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1767578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2461 }, { "epoch": 1.9696, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2462 }, { "epoch": 1.9704000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.23046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2463 }, { "epoch": 1.9712, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2464 }, { "epoch": 1.972, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2465 }, { "epoch": 1.9727999999999999, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.376953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2466 }, { "epoch": 1.9736, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2467 }, { "epoch": 1.9744000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2468 }, { "epoch": 1.9752, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.355224609375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2469 }, { "epoch": 1.976, "grad_norm": 0.1438661813735962, "learning_rate": 1.7066666666666667e-05, "loss": 1.6458984375, "step": 2470 }, { "epoch": 1.976, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2470 }, { "epoch": 1.9768, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2471 }, { "epoch": 1.9776, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2472 }, { "epoch": 1.9784000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2473 }, { "epoch": 1.9792, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2474 }, { "epoch": 1.98, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2475 }, { "epoch": 1.9808, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117919921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2476 }, { "epoch": 1.9816, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2477 }, { "epoch": 1.9824000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2478 }, { "epoch": 1.9832, "model.model.language_model.layers.26.mlp.down_proj": 0.19921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.12890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354736328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2479 }, { "epoch": 1.984, "grad_norm": 0.1578189581632614, "learning_rate": 1.6933333333333333e-05, "loss": 1.61044921875, "step": 2480 }, { "epoch": 1.984, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.23046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1884765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2480 }, { "epoch": 1.9848, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2481 }, { "epoch": 1.9856, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2482 }, { "epoch": 1.9864000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775634765625, "model.model.language_model.layers.26.mlp.up_proj": 0.1875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2483 }, { "epoch": 1.9872, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041748046875, "step": 2484 }, { "epoch": 1.988, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2485 }, { "epoch": 1.9888, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2486 }, { "epoch": 1.9896, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11767578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2487 }, { "epoch": 1.9904, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.354248046875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2488 }, { "epoch": 1.9912, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2489 }, { "epoch": 1.992, "grad_norm": 0.13348819315433502, "learning_rate": 1.6800000000000002e-05, "loss": 1.63896484375, "step": 2490 }, { "epoch": 1.992, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2490 }, { "epoch": 1.9928, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2491 }, { "epoch": 1.9936, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.117431640625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2492 }, { "epoch": 1.9944, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2493 }, { "epoch": 1.9952, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2494 }, { "epoch": 1.996, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2495 }, { "epoch": 1.9968, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2496 }, { "epoch": 1.9976, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35400390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2497 }, { "epoch": 1.9984, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.664306640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2498 }, { "epoch": 1.9992, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1787109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2499 }, { "epoch": 2.0, "grad_norm": 0.1513144075870514, "learning_rate": 1.6666666666666667e-05, "loss": 1.63203125, "step": 2500 }, { "epoch": 2.0, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405517578125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2500 }, { "epoch": 2.0008, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.23046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2501 }, { "epoch": 2.0016, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2502 }, { "epoch": 2.0024, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2503 }, { "epoch": 2.0032, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2504 }, { "epoch": 2.004, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353759765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2505 }, { "epoch": 2.0048, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2506 }, { "epoch": 2.0056, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2507 }, { "epoch": 2.0064, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775390625, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2508 }, { "epoch": 2.0072, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2509 }, { "epoch": 2.008, "grad_norm": 0.15850381553173065, "learning_rate": 1.6533333333333333e-05, "loss": 1.63291015625, "step": 2510 }, { "epoch": 2.008, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.23046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2510 }, { "epoch": 2.0088, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2511 }, { "epoch": 2.0096, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2512 }, { "epoch": 2.0104, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.1796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2513 }, { "epoch": 2.0112, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2514 }, { "epoch": 2.012, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1171875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2515 }, { "epoch": 2.0128, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2516 }, { "epoch": 2.0136, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2517 }, { "epoch": 2.0144, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2518 }, { "epoch": 2.0152, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2519 }, { "epoch": 2.016, "grad_norm": 0.1416153460741043, "learning_rate": 1.6400000000000002e-05, "loss": 1.63486328125, "step": 2520 }, { "epoch": 2.016, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2520 }, { "epoch": 2.0168, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2521 }, { "epoch": 2.0176, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04150390625, "step": 2522 }, { "epoch": 2.0184, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2523 }, { "epoch": 2.0192, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2524 }, { "epoch": 2.02, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2525 }, { "epoch": 2.0208, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2526 }, { "epoch": 2.0216, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6640625, "model.model.language_model.layers.26.self_attn.o_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2527 }, { "epoch": 2.0224, "model.model.language_model.layers.26.mlp.down_proj": 0.232421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.201171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.1650390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663818359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2528 }, { "epoch": 2.0232, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.775146484375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663818359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.833251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2529 }, { "epoch": 2.024, "grad_norm": 0.17257295548915863, "learning_rate": 1.6266666666666665e-05, "loss": 1.62333984375, "step": 2530 }, { "epoch": 2.024, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663818359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2530 }, { "epoch": 2.0248, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663818359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2531 }, { "epoch": 2.0256, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663818359375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2532 }, { "epoch": 2.0264, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2533 }, { "epoch": 2.0272, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2534 }, { "epoch": 2.028, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2535 }, { "epoch": 2.0288, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.116943359375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2536 }, { "epoch": 2.0296, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2537 }, { "epoch": 2.0304, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2538 }, { "epoch": 2.0312, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2539 }, { "epoch": 2.032, "grad_norm": 0.15301206707954407, "learning_rate": 1.6133333333333334e-05, "loss": 1.67734375, "step": 2540 }, { "epoch": 2.032, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2540 }, { "epoch": 2.0328, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.353271484375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2541 }, { "epoch": 2.0336, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66357421875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2542 }, { "epoch": 2.0344, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663330078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2543 }, { "epoch": 2.0352, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.405029296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663330078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2544 }, { "epoch": 2.036, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.663330078125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2545 }, { "epoch": 2.0368, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2546 }, { "epoch": 2.0376, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2547 }, { "epoch": 2.0384, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2548 }, { "epoch": 2.0392, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2549 }, { "epoch": 2.04, "grad_norm": 0.1464584320783615, "learning_rate": 1.6000000000000003e-05, "loss": 1.62080078125, "step": 2550 }, { "epoch": 2.04, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2550 }, { "epoch": 2.0408, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2551 }, { "epoch": 2.0416, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2552 }, { "epoch": 2.0424, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2553 }, { "epoch": 2.0432, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35302734375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2554 }, { "epoch": 2.044, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6630859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2555 }, { "epoch": 2.0448, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8330078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2556 }, { "epoch": 2.0456, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2557 }, { "epoch": 2.0464, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2558 }, { "epoch": 2.0472, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2559 }, { "epoch": 2.048, "grad_norm": 0.14394184947013855, "learning_rate": 1.586666666666667e-05, "loss": 1.67265625, "step": 2560 }, { "epoch": 2.048, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2560 }, { "epoch": 2.0488, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2561 }, { "epoch": 2.0496, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2562 }, { "epoch": 2.0504, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2563 }, { "epoch": 2.0512, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2564 }, { "epoch": 2.052, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2565 }, { "epoch": 2.0528, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2566 }, { "epoch": 2.0536, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2567 }, { "epoch": 2.0544, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2568 }, { "epoch": 2.0552, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2569 }, { "epoch": 2.056, "grad_norm": 0.14933611452579498, "learning_rate": 1.5733333333333334e-05, "loss": 1.630859375, "step": 2570 }, { "epoch": 2.056, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2570 }, { "epoch": 2.0568, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40478515625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2571 }, { "epoch": 2.0576, "model.model.language_model.layers.26.mlp.down_proj": 0.23828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832763671875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2572 }, { "epoch": 2.0584, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2573 }, { "epoch": 2.0592, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83251953125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2574 }, { "epoch": 2.06, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2575 }, { "epoch": 2.0608, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2576 }, { "epoch": 2.0616, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2577 }, { "epoch": 2.0624, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2578 }, { "epoch": 2.0632, "model.model.language_model.layers.26.mlp.down_proj": 0.236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11669921875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041259765625, "step": 2579 }, { "epoch": 2.064, "grad_norm": 0.12633879482746124, "learning_rate": 1.56e-05, "loss": 1.64130859375, "step": 2580 }, { "epoch": 2.064, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2580 }, { "epoch": 2.0648, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2581 }, { "epoch": 2.0656, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352783203125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2582 }, { "epoch": 2.0664, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2583 }, { "epoch": 2.0672, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2584 }, { "epoch": 2.068, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2585 }, { "epoch": 2.0688, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.279296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2586 }, { "epoch": 2.0696, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2587 }, { "epoch": 2.0704, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2588 }, { "epoch": 2.0712, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2589 }, { "epoch": 2.072, "grad_norm": 0.17006492614746094, "learning_rate": 1.546666666666667e-05, "loss": 1.6365234375, "step": 2590 }, { "epoch": 2.072, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662841796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2590 }, { "epoch": 2.0728, "model.model.language_model.layers.26.mlp.down_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2591 }, { "epoch": 2.0736, "model.model.language_model.layers.26.mlp.down_proj": 0.23828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2592 }, { "epoch": 2.0744, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3525390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2593 }, { "epoch": 2.0752, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2594 }, { "epoch": 2.076, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66259765625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2595 }, { "epoch": 2.0768, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2596 }, { "epoch": 2.0776, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2597 }, { "epoch": 2.0784, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2598 }, { "epoch": 2.0792, "model.model.language_model.layers.26.mlp.down_proj": 0.33984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2599 }, { "epoch": 2.08, "grad_norm": 0.13152460753917694, "learning_rate": 1.5333333333333334e-05, "loss": 1.6341796875, "step": 2600 }, { "epoch": 2.08, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77490234375, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2600 }, { "epoch": 2.0808, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2601 }, { "epoch": 2.0816, "model.model.language_model.layers.26.mlp.down_proj": 0.375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.34375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.345703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2602 }, { "epoch": 2.0824, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2603 }, { "epoch": 2.0832, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2604 }, { "epoch": 2.084, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2605 }, { "epoch": 2.0848, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2606 }, { "epoch": 2.0856, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.352294921875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2607 }, { "epoch": 2.0864, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.662109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2608 }, { "epoch": 2.0872, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2609 }, { "epoch": 2.088, "grad_norm": 0.1700315624475479, "learning_rate": 1.52e-05, "loss": 1.6314453125, "step": 2610 }, { "epoch": 2.088, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2610 }, { "epoch": 2.0888, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2611 }, { "epoch": 2.0896, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.361328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2612 }, { "epoch": 2.0904, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661865234375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2613 }, { "epoch": 2.0912, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66162109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35205078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2614 }, { "epoch": 2.092, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.661376953125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2615 }, { "epoch": 2.0928, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6611328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2616 }, { "epoch": 2.0936, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6611328125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2617 }, { "epoch": 2.0944, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660888671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2618 }, { "epoch": 2.0952, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660888671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2619 }, { "epoch": 2.096, "grad_norm": 0.1868623048067093, "learning_rate": 1.5066666666666668e-05, "loss": 1.62392578125, "step": 2620 }, { "epoch": 2.096, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660888671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2620 }, { "epoch": 2.0968, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1162109375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660888671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2621 }, { "epoch": 2.0976, "model.model.language_model.layers.26.mlp.down_proj": 0.341796875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660888671875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.041015625, "step": 2622 }, { "epoch": 2.0984, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2623 }, { "epoch": 2.0992, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.832275390625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2624 }, { "epoch": 2.1, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2625 }, { "epoch": 2.1008, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2626 }, { "epoch": 2.1016, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2627 }, { "epoch": 2.1024, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2628 }, { "epoch": 2.1032, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2629 }, { "epoch": 2.104, "grad_norm": 0.15823231637477875, "learning_rate": 1.4933333333333335e-05, "loss": 1.6294921875, "step": 2630 }, { "epoch": 2.104, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2630 }, { "epoch": 2.1048, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351806640625, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2631 }, { "epoch": 2.1056, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2632 }, { "epoch": 2.1064, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2633 }, { "epoch": 2.1072, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2634 }, { "epoch": 2.108, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66064453125, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2635 }, { "epoch": 2.1088, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2636 }, { "epoch": 2.1096, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2637 }, { "epoch": 2.1104, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2638 }, { "epoch": 2.1112, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2639 }, { "epoch": 2.112, "grad_norm": 0.17463034391403198, "learning_rate": 1.48e-05, "loss": 1.63935546875, "step": 2640 }, { "epoch": 2.112, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2640 }, { "epoch": 2.1128, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2641 }, { "epoch": 2.1136, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2642 }, { "epoch": 2.1144, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774658203125, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.351318359375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2643 }, { "epoch": 2.1152, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2644 }, { "epoch": 2.116, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404541015625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2645 }, { "epoch": 2.1168, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2646 }, { "epoch": 2.1176, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.660400390625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2647 }, { "epoch": 2.1184, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.66015625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2648 }, { "epoch": 2.1192, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2649 }, { "epoch": 2.12, "grad_norm": 0.1558074802160263, "learning_rate": 1.4666666666666668e-05, "loss": 1.62109375, "step": 2650 }, { "epoch": 2.12, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2650 }, { "epoch": 2.1208, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2651 }, { "epoch": 2.1216, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.30859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2652 }, { "epoch": 2.1224, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2653 }, { "epoch": 2.1232, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2654 }, { "epoch": 2.124, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404296875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2655 }, { "epoch": 2.1248, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2656 }, { "epoch": 2.1256, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2657 }, { "epoch": 2.1264, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35107421875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2658 }, { "epoch": 2.1272, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2659 }, { "epoch": 2.128, "grad_norm": 0.14637677371501923, "learning_rate": 1.4533333333333335e-05, "loss": 1.62587890625, "step": 2660 }, { "epoch": 2.128, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115966796875, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2660 }, { "epoch": 2.1288, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2661 }, { "epoch": 2.1296, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2662 }, { "epoch": 2.1304, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2663 }, { "epoch": 2.1312, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2664 }, { "epoch": 2.132, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2665 }, { "epoch": 2.1328, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2666 }, { "epoch": 2.1336, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.404052734375, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2667 }, { "epoch": 2.1344, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2668 }, { "epoch": 2.1352, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2669 }, { "epoch": 2.136, "grad_norm": 0.13508115708827972, "learning_rate": 1.44e-05, "loss": 1.61650390625, "step": 2670 }, { "epoch": 2.136, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2670 }, { "epoch": 2.1368, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2671 }, { "epoch": 2.1376, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2672 }, { "epoch": 2.1384, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2673 }, { "epoch": 2.1391999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.40234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40380859375, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2674 }, { "epoch": 2.14, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2675 }, { "epoch": 2.1408, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350830078125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2676 }, { "epoch": 2.1416, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2677 }, { "epoch": 2.1424, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2678 }, { "epoch": 2.1432, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83203125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2679 }, { "epoch": 2.144, "grad_norm": 0.1357879936695099, "learning_rate": 1.4266666666666667e-05, "loss": 1.63427734375, "step": 2680 }, { "epoch": 2.144, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7744140625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2680 }, { "epoch": 2.1448, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.13671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2681 }, { "epoch": 2.1456, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.30859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2682 }, { "epoch": 2.1464, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2683 }, { "epoch": 2.1471999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2684 }, { "epoch": 2.148, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2685 }, { "epoch": 2.1488, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2686 }, { "epoch": 2.1496, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.1845703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659912109375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2687 }, { "epoch": 2.1504, "model.model.language_model.layers.26.mlp.down_proj": 0.244140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2080078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2688 }, { "epoch": 2.1512000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2689 }, { "epoch": 2.152, "grad_norm": 0.13630124926567078, "learning_rate": 1.4133333333333334e-05, "loss": 1.643359375, "step": 2690 }, { "epoch": 2.152, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2690 }, { "epoch": 2.1528, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2691 }, { "epoch": 2.1536, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2692 }, { "epoch": 2.1544, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.774169921875, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831787109375, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2693 }, { "epoch": 2.1552, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11572265625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2694 }, { "epoch": 2.156, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2695 }, { "epoch": 2.1568, "model.model.language_model.layers.26.mlp.down_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2696 }, { "epoch": 2.1576, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2697 }, { "epoch": 2.1584, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2698 }, { "epoch": 2.1592000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2699 }, { "epoch": 2.16, "grad_norm": 0.15622420608997345, "learning_rate": 1.4000000000000001e-05, "loss": 1.63359375, "step": 2700 }, { "epoch": 2.16, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2700 }, { "epoch": 2.1608, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65966796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2701 }, { "epoch": 2.1616, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659423828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2702 }, { "epoch": 2.1624, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.659423828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2703 }, { "epoch": 2.1632, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6591796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2704 }, { "epoch": 2.164, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6591796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2705 }, { "epoch": 2.1648, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115478515625, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6591796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2706 }, { "epoch": 2.1656, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6591796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2707 }, { "epoch": 2.1664, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6591796875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2708 }, { "epoch": 2.1672, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2709 }, { "epoch": 2.168, "grad_norm": 0.1588020771741867, "learning_rate": 1.3866666666666667e-05, "loss": 1.60625, "step": 2710 }, { "epoch": 2.168, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2710 }, { "epoch": 2.1688, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2711 }, { "epoch": 2.1696, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1630859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3505859375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2712 }, { "epoch": 2.1704, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403564453125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2713 }, { "epoch": 2.1712, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4033203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2714 }, { "epoch": 2.172, "model.model.language_model.layers.26.mlp.down_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.1943359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4033203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2715 }, { "epoch": 2.1728, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2716 }, { "epoch": 2.1736, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2717 }, { "epoch": 2.1744, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2718 }, { "epoch": 2.1752, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77392578125, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.403076171875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2719 }, { "epoch": 2.176, "grad_norm": 0.1822715848684311, "learning_rate": 1.3733333333333335e-05, "loss": 1.5962890625, "step": 2720 }, { "epoch": 2.176, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040771484375, "step": 2720 }, { "epoch": 2.1768, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.22265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2721 }, { "epoch": 2.1776, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2722 }, { "epoch": 2.1784, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2723 }, { "epoch": 2.1792, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.130859375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2724 }, { "epoch": 2.18, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2725 }, { "epoch": 2.1808, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2726 }, { "epoch": 2.1816, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2727 }, { "epoch": 2.1824, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2728 }, { "epoch": 2.1832, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2729 }, { "epoch": 2.184, "grad_norm": 0.14385978877544403, "learning_rate": 1.3600000000000002e-05, "loss": 1.60205078125, "step": 2730 }, { "epoch": 2.184, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2730 }, { "epoch": 2.1848, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.115234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2731 }, { "epoch": 2.1856, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2732 }, { "epoch": 2.1864, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2733 }, { "epoch": 2.1872, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2734 }, { "epoch": 2.188, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2735 }, { "epoch": 2.1888, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2736 }, { "epoch": 2.1896, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2737 }, { "epoch": 2.1904, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2738 }, { "epoch": 2.1912, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2739 }, { "epoch": 2.192, "grad_norm": 0.1550496369600296, "learning_rate": 1.3466666666666666e-05, "loss": 1.62890625, "step": 2740 }, { "epoch": 2.192, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2740 }, { "epoch": 2.1928, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2741 }, { "epoch": 2.1936, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7734375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2742 }, { "epoch": 2.1944, "model.model.language_model.layers.26.mlp.down_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2743 }, { "epoch": 2.1952, "model.model.language_model.layers.26.mlp.down_proj": 0.33203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2744 }, { "epoch": 2.196, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2745 }, { "epoch": 2.1968, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2746 }, { "epoch": 2.1976, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2747 }, { "epoch": 2.1984, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2748 }, { "epoch": 2.1992, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2749 }, { "epoch": 2.2, "grad_norm": 0.14937640726566315, "learning_rate": 1.3333333333333333e-05, "loss": 1.60224609375, "step": 2750 }, { "epoch": 2.2, "model.model.language_model.layers.26.mlp.down_proj": 0.251953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2750 }, { "epoch": 2.2008, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2751 }, { "epoch": 2.2016, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2752 }, { "epoch": 2.2024, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2753 }, { "epoch": 2.2032, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2754 }, { "epoch": 2.204, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2755 }, { "epoch": 2.2048, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2756 }, { "epoch": 2.2056, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2757 }, { "epoch": 2.2064, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2758 }, { "epoch": 2.2072, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.212890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2759 }, { "epoch": 2.208, "grad_norm": 0.17556296288967133, "learning_rate": 1.32e-05, "loss": 1.6279296875, "step": 2760 }, { "epoch": 2.208, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2760 }, { "epoch": 2.2088, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2761 }, { "epoch": 2.2096, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658935546875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2762 }, { "epoch": 2.2104, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2763 }, { "epoch": 2.2112, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.350341796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2764 }, { "epoch": 2.212, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2765 }, { "epoch": 2.2128, "model.model.language_model.layers.26.mlp.down_proj": 0.376953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.34765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2766 }, { "epoch": 2.2136, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2767 }, { "epoch": 2.2144, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2768 }, { "epoch": 2.2152, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.35009765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2769 }, { "epoch": 2.216, "grad_norm": 0.14851361513137817, "learning_rate": 1.3066666666666666e-05, "loss": 1.64072265625, "step": 2770 }, { "epoch": 2.216, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2770 }, { "epoch": 2.2168, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2771 }, { "epoch": 2.2176, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65869140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2772 }, { "epoch": 2.2184, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1826171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2773 }, { "epoch": 2.2192, "model.model.language_model.layers.26.mlp.down_proj": 0.208984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.224609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.12109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2774 }, { "epoch": 2.22, "model.model.language_model.layers.26.mlp.down_proj": 0.3203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2775 }, { "epoch": 2.2208, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.04052734375, "step": 2776 }, { "epoch": 2.2216, "model.model.language_model.layers.26.mlp.down_proj": 0.333984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2777 }, { "epoch": 2.2224, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2778 }, { "epoch": 2.2232, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2779 }, { "epoch": 2.224, "grad_norm": 0.15270575881004333, "learning_rate": 1.2933333333333334e-05, "loss": 1.62568359375, "step": 2780 }, { "epoch": 2.224, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2780 }, { "epoch": 2.2248, "model.model.language_model.layers.26.mlp.down_proj": 0.232421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349853515625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2781 }, { "epoch": 2.2256, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2782 }, { "epoch": 2.2264, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.31640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2783 }, { "epoch": 2.2272, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.349365234375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2784 }, { "epoch": 2.228, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2785 }, { "epoch": 2.2288, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2786 }, { "epoch": 2.2296, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2787 }, { "epoch": 2.2304, "model.model.language_model.layers.26.mlp.down_proj": 0.3359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.287109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2788 }, { "epoch": 2.2312, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2789 }, { "epoch": 2.232, "grad_norm": 0.13928750157356262, "learning_rate": 1.2800000000000001e-05, "loss": 1.62333984375, "step": 2790 }, { "epoch": 2.232, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2790 }, { "epoch": 2.2328, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2791 }, { "epoch": 2.2336, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658447265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2792 }, { "epoch": 2.2344, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.773193359375, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2793 }, { "epoch": 2.2352, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.658203125, "model.model.language_model.layers.26.self_attn.o_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2794 }, { "epoch": 2.2359999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2795 }, { "epoch": 2.2368, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40283203125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2796 }, { "epoch": 2.2376, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2797 }, { "epoch": 2.2384, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2798 }, { "epoch": 2.2392, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.32421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.173828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2799 }, { "epoch": 2.24, "grad_norm": 0.17444953322410583, "learning_rate": 1.2666666666666668e-05, "loss": 1.6287109375, "step": 2800 }, { "epoch": 2.24, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2800 }, { "epoch": 2.2408, "model.model.language_model.layers.26.mlp.down_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402587890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2801 }, { "epoch": 2.2416, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2802 }, { "epoch": 2.2424, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2803 }, { "epoch": 2.2432, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2804 }, { "epoch": 2.2439999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.26171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.322265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2805 }, { "epoch": 2.2448, "model.model.language_model.layers.26.mlp.down_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.12890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2806 }, { "epoch": 2.2456, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77294921875, "model.model.language_model.layers.26.mlp.up_proj": 0.248046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2807 }, { "epoch": 2.2464, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772705078125, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2808 }, { "epoch": 2.2472, "model.model.language_model.layers.26.mlp.down_proj": 0.318359375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772705078125, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2809 }, { "epoch": 2.248, "grad_norm": 0.16816987097263336, "learning_rate": 1.2533333333333332e-05, "loss": 1.611328125, "step": 2810 }, { "epoch": 2.248, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772705078125, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2810 }, { "epoch": 2.2488, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772705078125, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2811 }, { "epoch": 2.2496, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2812 }, { "epoch": 2.2504, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.220703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2813 }, { "epoch": 2.2512, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2814 }, { "epoch": 2.252, "model.model.language_model.layers.26.mlp.down_proj": 0.2421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2815 }, { "epoch": 2.2528, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2816 }, { "epoch": 2.2536, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2817 }, { "epoch": 2.2544, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2818 }, { "epoch": 2.2552, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2819 }, { "epoch": 2.2560000000000002, "grad_norm": 0.17005501687526703, "learning_rate": 1.24e-05, "loss": 1.5787109375, "step": 2820 }, { "epoch": 2.2560000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.7724609375, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2490234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2820 }, { "epoch": 2.2568, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.17578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2821 }, { "epoch": 2.2576, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2822 }, { "epoch": 2.2584, "model.model.language_model.layers.26.mlp.down_proj": 0.248046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.193359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2823 }, { "epoch": 2.2592, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2824 }, { "epoch": 2.26, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2825 }, { "epoch": 2.2608, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2826 }, { "epoch": 2.2616, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.20703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2827 }, { "epoch": 2.2624, "model.model.language_model.layers.26.mlp.down_proj": 0.259765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2828 }, { "epoch": 2.2632, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2001953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34912109375, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2829 }, { "epoch": 2.2640000000000002, "grad_norm": 0.16096945106983185, "learning_rate": 1.2266666666666667e-05, "loss": 1.59326171875, "step": 2830 }, { "epoch": 2.2640000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2830 }, { "epoch": 2.2648, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2831 }, { "epoch": 2.2656, "model.model.language_model.layers.26.mlp.down_proj": 0.2109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.1767578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2832 }, { "epoch": 2.2664, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2833 }, { "epoch": 2.2672, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2834 }, { "epoch": 2.268, "model.model.language_model.layers.26.mlp.down_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40234375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2835 }, { "epoch": 2.2688, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657958984375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2836 }, { "epoch": 2.2696, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2837 }, { "epoch": 2.2704, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2838 }, { "epoch": 2.2712, "model.model.language_model.layers.26.mlp.down_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.1845703125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1318359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2839 }, { "epoch": 2.2720000000000002, "grad_norm": 0.13710160553455353, "learning_rate": 1.2133333333333335e-05, "loss": 1.64072265625, "step": 2840 }, { "epoch": 2.2720000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.271484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2840 }, { "epoch": 2.2728, "model.model.language_model.layers.26.mlp.down_proj": 0.37109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.384765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.772216796875, "model.model.language_model.layers.26.mlp.up_proj": 0.33203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.402099609375, "model.model.language_model.layers.26.self_attn.k_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.3515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.310546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2841 }, { "epoch": 2.2736, "model.model.language_model.layers.26.mlp.down_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1884765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2842 }, { "epoch": 2.2744, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2843 }, { "epoch": 2.2752, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2844 }, { "epoch": 2.276, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2845 }, { "epoch": 2.2768, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2846 }, { "epoch": 2.2776, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2847 }, { "epoch": 2.2784, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2848 }, { "epoch": 2.2792, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1650390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2849 }, { "epoch": 2.2800000000000002, "grad_norm": 0.15144672989845276, "learning_rate": 1.2e-05, "loss": 1.592578125, "step": 2850 }, { "epoch": 2.2800000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2850 }, { "epoch": 2.2808, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2851 }, { "epoch": 2.2816, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2852 }, { "epoch": 2.2824, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.306640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2853 }, { "epoch": 2.2832, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2854 }, { "epoch": 2.284, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348876953125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2855 }, { "epoch": 2.2848, "model.model.language_model.layers.26.mlp.down_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.228515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1728515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.12451171875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2856 }, { "epoch": 2.2856, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2857 }, { "epoch": 2.2864, "model.model.language_model.layers.26.mlp.down_proj": 0.330078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.263671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40185546875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65771484375, "model.model.language_model.layers.26.self_attn.o_proj": 0.306640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2858 }, { "epoch": 2.2872, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2859 }, { "epoch": 2.288, "grad_norm": 0.17804458737373352, "learning_rate": 1.1866666666666668e-05, "loss": 1.62626953125, "step": 2860 }, { "epoch": 2.288, "model.model.language_model.layers.26.mlp.down_proj": 0.3046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2860 }, { "epoch": 2.2888, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2861 }, { "epoch": 2.2896, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2862 }, { "epoch": 2.2904, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2863 }, { "epoch": 2.2912, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2864 }, { "epoch": 2.292, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114990234375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.30078125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2865 }, { "epoch": 2.2928, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2866 }, { "epoch": 2.2936, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2392578125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2867 }, { "epoch": 2.2944, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2868 }, { "epoch": 2.2952, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2869 }, { "epoch": 2.296, "grad_norm": 0.18305891752243042, "learning_rate": 1.1733333333333333e-05, "loss": 1.62626953125, "step": 2870 }, { "epoch": 2.296, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2870 }, { "epoch": 2.2968, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2060546875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2871 }, { "epoch": 2.2976, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2872 }, { "epoch": 2.2984, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2873 }, { "epoch": 2.2992, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2874 }, { "epoch": 2.3, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3486328125, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83154296875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2875 }, { "epoch": 2.3008, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.244140625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2876 }, { "epoch": 2.3016, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2877 }, { "epoch": 2.3024, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2878 }, { "epoch": 2.3032, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2879 }, { "epoch": 2.304, "grad_norm": 0.14519572257995605, "learning_rate": 1.16e-05, "loss": 1.61455078125, "step": 2880 }, { "epoch": 2.304, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.25, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2880 }, { "epoch": 2.3048, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2158203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2881 }, { "epoch": 2.3056, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.348388671875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2882 }, { "epoch": 2.3064, "model.model.language_model.layers.26.mlp.down_proj": 0.369140625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.26953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401611328125, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.3359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.240234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2883 }, { "epoch": 2.3072, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2884 }, { "epoch": 2.308, "model.model.language_model.layers.26.mlp.down_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.185546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1259765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2885 }, { "epoch": 2.3088, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2886 }, { "epoch": 2.3096, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2887 }, { "epoch": 2.3104, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.831298828125, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2888 }, { "epoch": 2.3112, "model.model.language_model.layers.26.mlp.down_proj": 0.34765625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.3203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2889 }, { "epoch": 2.312, "grad_norm": 0.16130973398685455, "learning_rate": 1.1466666666666666e-05, "loss": 1.61513671875, "step": 2890 }, { "epoch": 2.312, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2890 }, { "epoch": 2.3128, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34814453125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2891 }, { "epoch": 2.3136, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2892 }, { "epoch": 2.3144, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2893 }, { "epoch": 2.3152, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2894 }, { "epoch": 2.316, "model.model.language_model.layers.26.mlp.down_proj": 0.23046875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1298828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.19140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2895 }, { "epoch": 2.3168, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2896 }, { "epoch": 2.3176, "model.model.language_model.layers.26.mlp.down_proj": 0.326171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2897 }, { "epoch": 2.3184, "model.model.language_model.layers.26.mlp.down_proj": 0.3515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.318359375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.27734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.25, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2898 }, { "epoch": 2.3192, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2899 }, { "epoch": 2.32, "grad_norm": 0.13477030396461487, "learning_rate": 1.1333333333333334e-05, "loss": 1.616796875, "step": 2900 }, { "epoch": 2.32, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2900 }, { "epoch": 2.3208, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.657470703125, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2901 }, { "epoch": 2.3216, "model.model.language_model.layers.26.mlp.down_proj": 0.240234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2902 }, { "epoch": 2.3224, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.11474609375, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2903 }, { "epoch": 2.3232, "model.model.language_model.layers.26.mlp.down_proj": 0.25, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2904 }, { "epoch": 2.324, "model.model.language_model.layers.26.mlp.down_proj": 0.357421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.28515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.3203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.28515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2905 }, { "epoch": 2.3247999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.275390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2906 }, { "epoch": 2.3256, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.347900390625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2907 }, { "epoch": 2.3264, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.291015625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2908 }, { "epoch": 2.3272, "model.model.language_model.layers.26.mlp.down_proj": 0.302734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.162109375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34765625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.8310546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2909 }, { "epoch": 2.328, "grad_norm": 0.12989327311515808, "learning_rate": 1.1200000000000001e-05, "loss": 1.598046875, "step": 2910 }, { "epoch": 2.328, "model.model.language_model.layers.26.mlp.down_proj": 0.23828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.12890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34716796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.181640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2910 }, { "epoch": 2.3288, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2431640625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34716796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2911 }, { "epoch": 2.3296, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.267578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34716796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2912 }, { "epoch": 2.3304, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.6572265625, "model.model.language_model.layers.26.self_attn.o_proj": 0.21875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2913 }, { "epoch": 2.3312, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2914 }, { "epoch": 2.332, "model.model.language_model.layers.26.mlp.down_proj": 0.27734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.21875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2915 }, { "epoch": 2.3327999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.271484375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2916 }, { "epoch": 2.3336, "model.model.language_model.layers.26.mlp.down_proj": 0.2470703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2917 }, { "epoch": 2.3344, "model.model.language_model.layers.26.mlp.down_proj": 0.208984375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1806640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1279296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.1806640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2918 }, { "epoch": 2.3352, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.248046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.205078125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2919 }, { "epoch": 2.336, "grad_norm": 0.16598162055015564, "learning_rate": 1.1066666666666667e-05, "loss": 1.5888671875, "step": 2920 }, { "epoch": 2.336, "model.model.language_model.layers.26.mlp.down_proj": 0.30078125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2920 }, { "epoch": 2.3368, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.240234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2921 }, { "epoch": 2.3376, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2922 }, { "epoch": 2.3384, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.29296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2923 }, { "epoch": 2.3392, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.259765625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1376953125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.197265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2924 }, { "epoch": 2.34, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.25390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1591796875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.318359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830810546875, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2925 }, { "epoch": 2.3407999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.25390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.4013671875, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.224609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2926 }, { "epoch": 2.3416, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28515625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2275390625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2216796875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2927 }, { "epoch": 2.3424, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2928 }, { "epoch": 2.3432, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2929 }, { "epoch": 2.344, "grad_norm": 0.13445378839969635, "learning_rate": 1.0933333333333334e-05, "loss": 1.62109375, "step": 2930 }, { "epoch": 2.344, "model.model.language_model.layers.26.mlp.down_proj": 0.24609375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1884765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.401123046875, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1884765625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2930 }, { "epoch": 2.3448, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2490234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65673828125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2931 }, { "epoch": 2.3456, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.040283203125, "step": 2932 }, { "epoch": 2.3464, "model.model.language_model.layers.26.mlp.down_proj": 0.314453125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2933 }, { "epoch": 2.3472, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.294921875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2934 }, { "epoch": 2.348, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.142578125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2119140625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2935 }, { "epoch": 2.3487999999999998, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2936 }, { "epoch": 2.3496, "model.model.language_model.layers.26.mlp.down_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1962890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2937 }, { "epoch": 2.3504, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2294921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2938 }, { "epoch": 2.3512, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656494140625, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2939 }, { "epoch": 2.352, "grad_norm": 0.14417335391044617, "learning_rate": 1.08e-05, "loss": 1.6091796875, "step": 2940 }, { "epoch": 2.352, "model.model.language_model.layers.26.mlp.down_proj": 0.2890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65625, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2940 }, { "epoch": 2.3528000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.251953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1318359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65625, "model.model.language_model.layers.26.self_attn.o_proj": 0.251953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.193359375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2941 }, { "epoch": 2.3536, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2197265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65625, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2109375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2942 }, { "epoch": 2.3544, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65625, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2943 }, { "epoch": 2.3552, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.244140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2944 }, { "epoch": 2.356, "model.model.language_model.layers.26.mlp.down_proj": 0.29296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.279296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1513671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346923828125, "model.model.language_model.layers.26.self_attn.q_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2945 }, { "epoch": 2.3568, "model.model.language_model.layers.26.mlp.down_proj": 0.267578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.14453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2946 }, { "epoch": 2.3576, "model.model.language_model.layers.26.mlp.down_proj": 0.337890625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.322265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.267578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1728515625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2947 }, { "epoch": 2.3584, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.302734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.234375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2948 }, { "epoch": 2.3592, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2333984375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2949 }, { "epoch": 2.36, "grad_norm": 0.165225088596344, "learning_rate": 1.0666666666666667e-05, "loss": 1.61357421875, "step": 2950 }, { "epoch": 2.36, "model.model.language_model.layers.26.mlp.down_proj": 0.1982421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.114501953125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1640625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.11474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1640625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2950 }, { "epoch": 2.3608000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.23828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1904296875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2951 }, { "epoch": 2.3616, "model.model.language_model.layers.26.mlp.down_proj": 0.3125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1689453125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2952 }, { "epoch": 2.3624, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.21484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.83056640625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2953 }, { "epoch": 2.3632, "model.model.language_model.layers.26.mlp.down_proj": 0.2451171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1904296875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.185546875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.0400390625, "step": 2954 }, { "epoch": 2.364, "model.model.language_model.layers.26.mlp.down_proj": 0.23828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2955 }, { "epoch": 2.3648, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.232421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1533203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2956 }, { "epoch": 2.3656, "model.model.language_model.layers.26.mlp.down_proj": 0.263671875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.255859375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2021484375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2957 }, { "epoch": 2.3664, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.205078125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2958 }, { "epoch": 2.3672, "model.model.language_model.layers.26.mlp.down_proj": 0.287109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.201171875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.158203125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2959 }, { "epoch": 2.368, "grad_norm": 0.15476389229297638, "learning_rate": 1.0533333333333335e-05, "loss": 1.5845703125, "step": 2960 }, { "epoch": 2.368, "model.model.language_model.layers.26.mlp.down_proj": 0.26171875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1416015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.208984375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2960 }, { "epoch": 2.3688000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.2353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2421875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.19140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.169921875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.1962890625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2021484375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2961 }, { "epoch": 2.3696, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2333984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2962 }, { "epoch": 2.3704, "model.model.language_model.layers.26.mlp.down_proj": 0.306640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.228515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2963 }, { "epoch": 2.3712, "model.model.language_model.layers.26.mlp.down_proj": 0.322265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.3046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.251953125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1552734375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.291015625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2964 }, { "epoch": 2.372, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.212890625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.656005859375, "model.model.language_model.layers.26.self_attn.o_proj": 0.234375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.3466796875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2965 }, { "epoch": 2.3728, "model.model.language_model.layers.26.mlp.down_proj": 0.30859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.298828125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.16015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2314453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2966 }, { "epoch": 2.3736, "model.model.language_model.layers.26.mlp.down_proj": 0.32421875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.296875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.234375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2967 }, { "epoch": 2.3744, "model.model.language_model.layers.26.mlp.down_proj": 0.298828125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2890625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2421875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1455078125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.259765625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2080078125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2412109375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2968 }, { "epoch": 2.3752, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.24609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.271484375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2353515625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2969 }, { "epoch": 2.376, "grad_norm": 0.1447850912809372, "learning_rate": 1.04e-05, "loss": 1.59755859375, "step": 2970 }, { "epoch": 2.376, "model.model.language_model.layers.26.mlp.down_proj": 0.2177734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.18359375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1298828125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.65576171875, "model.model.language_model.layers.26.self_attn.o_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2970 }, { "epoch": 2.3768000000000002, "model.model.language_model.layers.26.mlp.down_proj": 0.34375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.330078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.283203125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1748046875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.283203125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.25, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.27734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2971 }, { "epoch": 2.3776, "model.model.language_model.layers.26.mlp.down_proj": 0.283203125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.24609375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.208984375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1474609375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2972 }, { "epoch": 2.3784, "model.model.language_model.layers.26.mlp.down_proj": 0.296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2578125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2099609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2451171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2973 }, { "epoch": 2.3792, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.267578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2974 }, { "epoch": 2.38, "model.model.language_model.layers.26.mlp.down_proj": 0.2431640625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.240234375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1357421875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.346435546875, "model.model.language_model.layers.26.self_attn.q_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.18359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2975 }, { "epoch": 2.3808, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2236328125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.1923828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1611328125, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.1845703125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2976 }, { "epoch": 2.3816, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.271484375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.77197265625, "model.model.language_model.layers.26.mlp.up_proj": 0.2119140625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26171875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.20703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2977 }, { "epoch": 2.3824, "model.model.language_model.layers.26.mlp.down_proj": 0.345703125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.314453125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.259765625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1669921875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.298828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2978 }, { "epoch": 2.3832, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2979 }, { "epoch": 2.384, "grad_norm": 0.1279883086681366, "learning_rate": 1.0266666666666668e-05, "loss": 1.5876953125, "step": 2980 }, { "epoch": 2.384, "model.model.language_model.layers.26.mlp.down_proj": 0.2392578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2373046875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.1865234375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1318359375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.1923828125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2980 }, { "epoch": 2.3848, "model.model.language_model.layers.26.mlp.down_proj": 0.353515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.310546875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1708984375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.29296875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.24609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.248046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2981 }, { "epoch": 2.3856, "model.model.language_model.layers.26.mlp.down_proj": 0.365234375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.33984375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.291015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1865234375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.328125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.294921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2982 }, { "epoch": 2.3864, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.228515625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2983 }, { "epoch": 2.3872, "model.model.language_model.layers.26.mlp.down_proj": 0.2578125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.263671875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2060546875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1435546875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2373046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.34619140625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2041015625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.203125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2984 }, { "epoch": 2.388, "model.model.language_model.layers.26.mlp.down_proj": 0.310546875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.30078125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.15625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.26953125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2985 }, { "epoch": 2.3888, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2255859375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.146484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.25390625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2986 }, { "epoch": 2.3896, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26953125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2138671875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1494140625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.216796875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2177734375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2987 }, { "epoch": 2.3904, "model.model.language_model.layers.26.mlp.down_proj": 0.275390625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23046875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2988 }, { "epoch": 2.3912, "model.model.language_model.layers.26.mlp.down_proj": 0.291015625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.287109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.23828125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.263671875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2197265625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2989 }, { "epoch": 2.392, "grad_norm": 0.1518896073102951, "learning_rate": 1.0133333333333333e-05, "loss": 1.5955078125, "step": 2990 }, { "epoch": 2.392, "model.model.language_model.layers.26.mlp.down_proj": 0.265625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.26171875, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2041015625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.138671875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.201171875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2990 }, { "epoch": 2.3928, "model.model.language_model.layers.26.mlp.down_proj": 0.28515625, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.224609375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.265625, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2099609375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.22265625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2991 }, { "epoch": 2.3936, "model.model.language_model.layers.26.mlp.down_proj": 0.279296875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.22265625, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.150390625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2158203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.2275390625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2992 }, { "epoch": 2.3944, "model.model.language_model.layers.26.mlp.down_proj": 0.2734375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.27734375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.216796875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.23828125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.2255859375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830322265625, "model.model.language_model.layers.26.self_attn.v_proj": 0.212890625, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2993 }, { "epoch": 2.3952, "model.model.language_model.layers.26.mlp.down_proj": 0.328125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.31640625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2578125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.166015625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.302734375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345947265625, "model.model.language_model.layers.26.self_attn.q_proj": 0.244140625, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2470703125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2994 }, { "epoch": 2.396, "model.model.language_model.layers.26.mlp.down_proj": 0.28125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.283203125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2314453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1572265625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.255859375, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.220703125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2995 }, { "epoch": 2.3968, "model.model.language_model.layers.26.mlp.down_proj": 0.2109375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.2412109375, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.189453125, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.134765625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.232421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.189453125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.1943359375, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2996 }, { "epoch": 2.3976, "model.model.language_model.layers.26.mlp.down_proj": 0.255859375, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.25390625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.19921875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1337890625, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.1982421875, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.19921875, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2001953125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2997 }, { "epoch": 2.3984, "model.model.language_model.layers.26.mlp.down_proj": 0.294921875, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.28125, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.23046875, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.154296875, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.28125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.21484375, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2236328125, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2998 }, { "epoch": 2.3992, "model.model.language_model.layers.26.mlp.down_proj": 0.26953125, "model.model.language_model.layers.26.mlp.down_proj/dead_latent_percentage": 0.1142578125, "model.model.language_model.layers.26.mlp.gate_proj": 0.265625, "model.model.language_model.layers.26.mlp.gate_proj/dead_latent_percentage": 0.771728515625, "model.model.language_model.layers.26.mlp.up_proj": 0.2109375, "model.model.language_model.layers.26.mlp.up_proj/dead_latent_percentage": 0.40087890625, "model.model.language_model.layers.26.self_attn.k_proj": 0.1396484375, "model.model.language_model.layers.26.self_attn.k_proj/dead_latent_percentage": 0.655517578125, "model.model.language_model.layers.26.self_attn.o_proj": 0.2578125, "model.model.language_model.layers.26.self_attn.o_proj/dead_latent_percentage": 0.345703125, "model.model.language_model.layers.26.self_attn.q_proj": 0.203125, "model.model.language_model.layers.26.self_attn.q_proj/dead_latent_percentage": 0.830078125, "model.model.language_model.layers.26.self_attn.v_proj": 0.2138671875, "model.model.language_model.layers.26.self_attn.v_proj/dead_latent_percentage": 0.039794921875, "step": 2999 }, { "epoch": 2.4, "grad_norm": 0.12872867286205292, "learning_rate": 1e-05, "loss": 1.575, "step": 3000 } ], "logging_steps": 10, "max_steps": 3750, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 8.634805589051965e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }