{ "auto_mapping": null, "base_model_name_or_path": "checkpoints/Uni-MoE-speech-base", "bias": "none", "fan_in_fan_out": false, "inference_mode": true, "init_lora_weights": true, "layers_pattern": null, "layers_to_transform": null, "lora_alpha": 16, "lora_dropout": 0.05, "modules_to_save": null, "peft_type": "LORA", "r": 8, "revision": null, "target_modules": [ "model.layers.7.mlp.experts.3.gate_proj", "model.layers.29.mlp.experts.3.up_proj", "model.layers.28.mlp.experts.2.up_proj", "model.layers.8.self_attn.q_proj", "model.layers.11.self_attn.o_proj", "model.layers.26.mlp.experts.1.down_proj", "model.layers.27.mlp.experts.3.gate_proj", "model.layers.26.mlp.experts.0.down_proj", "model.layers.31.mlp.experts.2.down_proj", "model.layers.27.mlp.experts.3.down_proj", "model.layers.23.self_attn.v_proj", "model.layers.8.self_attn.o_proj", "model.layers.28.self_attn.v_proj", "model.layers.22.mlp.experts.0.up_proj", "model.layers.19.self_attn.v_proj", "model.layers.28.mlp.experts.0.up_proj", "model.layers.25.self_attn.k_proj", "model.layers.10.mlp.experts.2.up_proj", "model.layers.23.self_attn.k_proj", "model.layers.17.self_attn.q_proj", "model.layers.19.mlp.experts.1.down_proj", "model.layers.11.self_attn.v_proj", "model.layers.16.self_attn.k_proj", "model.layers.8.mlp.experts.2.up_proj", "model.layers.4.self_attn.k_proj", "model.layers.31.mlp.experts.2.gate_proj", "model.layers.21.mlp.experts.3.up_proj", "model.layers.11.mlp.experts.2.gate_proj", "model.layers.23.mlp.experts.3.gate_proj", "model.layers.14.mlp.experts.3.gate_proj", "model.layers.18.mlp.experts.1.gate_proj", "model.layers.24.mlp.experts.3.up_proj", "model.layers.4.self_attn.q_proj", "model.layers.26.mlp.experts.3.gate_proj", "model.layers.20.mlp.experts.0.up_proj", "model.layers.5.mlp.experts.3.up_proj", "model.layers.21.mlp.experts.0.up_proj", "model.layers.3.self_attn.q_proj", "model.layers.27.mlp.experts.0.down_proj", "model.layers.12.self_attn.o_proj", "model.layers.28.self_attn.k_proj", "model.layers.5.mlp.experts.1.down_proj", "model.layers.7.mlp.experts.1.gate_proj", "model.layers.24.self_attn.k_proj", "model.layers.4.self_attn.o_proj", "model.layers.4.mlp.experts.0.down_proj", "model.layers.22.self_attn.v_proj", "model.layers.22.self_attn.o_proj", "model.layers.15.self_attn.k_proj", "model.layers.19.mlp.experts.3.up_proj", "model.layers.26.self_attn.k_proj", "model.layers.10.mlp.experts.1.down_proj", "model.layers.27.mlp.experts.0.up_proj", "model.layers.11.mlp.experts.3.gate_proj", "model.layers.30.mlp.experts.0.gate_proj", "model.layers.25.mlp.experts.0.gate_proj", "model.layers.26.self_attn.o_proj", "model.layers.9.mlp.experts.2.up_proj", "model.layers.18.mlp.experts.2.gate_proj", "model.layers.14.mlp.experts.1.up_proj", "model.layers.28.mlp.experts.3.down_proj", "model.layers.11.mlp.experts.1.down_proj", "model.layers.12.mlp.experts.2.down_proj", "model.layers.16.mlp.experts.3.up_proj", "model.layers.17.mlp.experts.0.down_proj", "model.layers.6.self_attn.v_proj", "model.layers.15.mlp.experts.2.down_proj", "model.layers.26.self_attn.v_proj", "model.layers.6.self_attn.o_proj", "model.layers.0.mlp.experts.1.up_proj", "model.layers.14.mlp.experts.0.up_proj", "model.layers.16.mlp.experts.1.up_proj", "model.layers.3.mlp.experts.1.down_proj", "model.layers.19.self_attn.q_proj", "model.layers.27.mlp.experts.2.up_proj", "model.layers.19.mlp.experts.0.gate_proj", "model.layers.5.self_attn.v_proj", "model.layers.2.self_attn.q_proj", "model.layers.9.mlp.experts.3.down_proj", "model.layers.21.mlp.experts.1.up_proj", "model.layers.4.mlp.experts.2.down_proj", "model.layers.3.mlp.experts.1.gate_proj", "model.layers.5.mlp.experts.1.up_proj", "model.layers.0.mlp.experts.3.gate_proj", "model.layers.28.mlp.experts.2.down_proj", "model.layers.6.mlp.experts.3.down_proj", "model.layers.11.mlp.experts.3.down_proj", "model.layers.30.mlp.experts.1.down_proj", "model.layers.11.mlp.experts.2.down_proj", "model.layers.18.mlp.experts.3.gate_proj", "model.layers.16.mlp.experts.2.gate_proj", "model.layers.1.self_attn.k_proj", "model.layers.22.mlp.experts.2.up_proj", "model.layers.9.mlp.experts.3.gate_proj", "model.layers.15.mlp.experts.3.up_proj", "model.layers.7.self_attn.v_proj", "model.layers.22.self_attn.q_proj", "model.layers.22.mlp.experts.2.gate_proj", "model.layers.30.mlp.experts.3.down_proj", "model.layers.28.mlp.experts.1.down_proj", "model.layers.5.mlp.experts.3.down_proj", "model.layers.9.self_attn.q_proj", "model.layers.17.mlp.experts.1.gate_proj", "model.layers.22.mlp.experts.3.down_proj", "model.layers.29.self_attn.v_proj", "model.layers.8.mlp.experts.3.gate_proj", "model.layers.29.self_attn.q_proj", "model.layers.4.self_attn.v_proj", "model.layers.6.mlp.experts.1.down_proj", "model.layers.28.mlp.experts.0.gate_proj", "model.layers.30.mlp.experts.3.up_proj", "model.layers.4.mlp.experts.0.up_proj", "model.layers.14.self_attn.q_proj", "model.layers.18.mlp.experts.1.down_proj", "model.layers.10.mlp.experts.0.up_proj", "model.layers.13.self_attn.k_proj", "model.layers.12.self_attn.v_proj", "model.layers.14.mlp.experts.3.up_proj", "model.layers.27.self_attn.q_proj", "model.layers.25.self_attn.q_proj", "model.layers.6.self_attn.q_proj", "model.layers.8.self_attn.v_proj", "model.layers.14.mlp.experts.0.down_proj", "model.layers.14.mlp.experts.2.up_proj", "model.layers.9.self_attn.v_proj", "model.layers.6.mlp.experts.1.gate_proj", "model.layers.13.self_attn.q_proj", "model.layers.6.mlp.experts.0.up_proj", "model.layers.30.self_attn.q_proj", "model.layers.18.self_attn.o_proj", "model.layers.2.mlp.experts.2.gate_proj", "model.layers.30.mlp.experts.2.gate_proj", "model.layers.3.mlp.experts.3.up_proj", "model.layers.10.mlp.experts.1.gate_proj", "model.layers.19.mlp.experts.0.up_proj", "model.layers.12.mlp.experts.1.up_proj", "model.layers.20.mlp.experts.2.up_proj", "model.layers.22.mlp.experts.1.up_proj", "model.layers.3.mlp.experts.0.down_proj", "model.layers.1.mlp.experts.1.up_proj", "model.layers.13.mlp.experts.2.gate_proj", "model.layers.10.self_attn.q_proj", "model.layers.26.mlp.experts.1.gate_proj", "model.layers.14.mlp.experts.3.down_proj", "model.layers.1.self_attn.o_proj", "model.layers.29.self_attn.o_proj", "model.layers.14.mlp.experts.0.gate_proj", "model.layers.10.mlp.experts.0.gate_proj", "model.layers.4.mlp.experts.1.up_proj", "model.layers.4.mlp.experts.2.gate_proj", "model.layers.20.mlp.experts.1.gate_proj", "model.layers.20.mlp.experts.1.up_proj", "model.layers.23.mlp.experts.2.up_proj", "model.layers.0.mlp.experts.0.gate_proj", "model.layers.15.mlp.experts.1.up_proj", "model.layers.10.mlp.experts.3.up_proj", "model.layers.5.mlp.experts.1.gate_proj", "model.layers.12.mlp.experts.0.gate_proj", "model.layers.8.mlp.experts.3.up_proj", "model.layers.4.mlp.experts.1.gate_proj", "model.layers.9.mlp.experts.0.down_proj", "model.layers.0.mlp.experts.0.up_proj", "model.layers.7.mlp.experts.2.up_proj", "model.layers.11.mlp.experts.2.up_proj", "model.layers.19.mlp.experts.0.down_proj", "model.layers.25.mlp.experts.2.gate_proj", "model.layers.24.mlp.experts.3.down_proj", "model.layers.10.mlp.experts.2.down_proj", "model.layers.20.self_attn.q_proj", "model.layers.29.mlp.experts.2.gate_proj", "model.layers.25.mlp.experts.1.up_proj", "model.layers.15.mlp.experts.1.down_proj", "model.layers.3.mlp.experts.1.up_proj", "model.layers.18.self_attn.k_proj", "model.layers.11.mlp.experts.0.gate_proj", "model.layers.21.mlp.experts.0.gate_proj", "model.layers.4.mlp.experts.3.gate_proj", "model.layers.23.mlp.experts.2.down_proj", "model.layers.5.mlp.experts.0.up_proj", "model.layers.24.mlp.experts.0.down_proj", "model.layers.24.mlp.experts.1.gate_proj", "model.layers.16.self_attn.o_proj", "model.layers.25.mlp.experts.2.down_proj", "model.layers.26.mlp.experts.0.up_proj", "model.layers.13.mlp.experts.1.gate_proj", "model.layers.22.mlp.experts.2.down_proj", "model.layers.11.mlp.experts.3.up_proj", "model.layers.20.mlp.experts.0.gate_proj", "model.layers.9.mlp.experts.0.gate_proj", "model.layers.1.mlp.experts.3.gate_proj", "model.layers.1.mlp.experts.3.down_proj", "model.layers.9.self_attn.o_proj", "model.layers.25.self_attn.o_proj", "model.layers.31.mlp.experts.2.up_proj", "model.layers.6.mlp.experts.2.down_proj", "model.layers.4.mlp.experts.1.down_proj", "model.layers.21.mlp.experts.2.gate_proj", "model.layers.27.mlp.experts.2.down_proj", "model.layers.19.mlp.experts.2.down_proj", "model.layers.10.mlp.experts.0.down_proj", "model.layers.10.self_attn.v_proj", "model.layers.11.mlp.experts.1.gate_proj", "model.layers.1.mlp.experts.2.gate_proj", "model.layers.16.mlp.experts.3.down_proj", "model.layers.31.mlp.experts.0.gate_proj", "model.layers.13.mlp.experts.3.down_proj", "model.layers.0.self_attn.k_proj", "model.layers.27.self_attn.o_proj", "model.layers.18.self_attn.v_proj", "model.layers.28.mlp.experts.0.down_proj", "model.layers.13.mlp.experts.0.down_proj", "model.layers.13.mlp.experts.0.up_proj", "model.layers.25.mlp.experts.1.down_proj", "model.layers.19.mlp.experts.1.up_proj", "model.layers.21.self_attn.v_proj", "model.layers.20.mlp.experts.1.down_proj", "model.layers.9.mlp.experts.1.up_proj", "model.layers.27.mlp.experts.1.up_proj", "model.layers.31.mlp.experts.1.down_proj", "model.layers.3.mlp.experts.3.down_proj", "model.layers.22.mlp.experts.1.down_proj", "model.layers.6.mlp.experts.1.up_proj", "model.layers.16.self_attn.q_proj", "model.layers.27.self_attn.k_proj", "model.layers.9.mlp.experts.0.up_proj", "model.layers.24.mlp.experts.0.gate_proj", "model.layers.20.self_attn.o_proj", "model.layers.23.mlp.experts.0.down_proj", "model.layers.29.mlp.experts.3.down_proj", "model.layers.5.mlp.experts.2.gate_proj", "model.layers.19.mlp.experts.3.down_proj", "model.layers.29.mlp.experts.0.gate_proj", "model.layers.9.mlp.experts.1.gate_proj", "model.layers.18.mlp.experts.2.up_proj", "model.layers.20.self_attn.k_proj", "model.layers.26.self_attn.q_proj", "model.layers.2.mlp.experts.1.up_proj", "model.layers.7.mlp.experts.3.up_proj", "model.layers.4.mlp.experts.0.gate_proj", "model.layers.5.mlp.experts.0.down_proj", "model.layers.12.self_attn.q_proj", "model.layers.7.mlp.experts.2.down_proj", "model.layers.30.mlp.experts.0.up_proj", "model.layers.0.self_attn.v_proj", "model.layers.22.mlp.experts.1.gate_proj", "model.layers.21.mlp.experts.3.down_proj", "model.layers.19.mlp.experts.2.up_proj", "model.layers.28.self_attn.o_proj", "model.layers.16.mlp.experts.2.down_proj", "model.layers.17.mlp.experts.3.down_proj", "model.layers.26.mlp.experts.0.gate_proj", "model.layers.12.mlp.experts.0.up_proj", "model.layers.29.mlp.experts.1.gate_proj", "model.layers.21.mlp.experts.2.down_proj", "model.layers.14.mlp.experts.2.down_proj", "model.layers.25.mlp.experts.0.up_proj", "model.layers.18.self_attn.q_proj", "model.layers.30.mlp.experts.3.gate_proj", "model.layers.21.mlp.experts.1.gate_proj", "model.layers.27.mlp.experts.2.gate_proj", "model.layers.17.mlp.experts.0.gate_proj", "model.layers.18.mlp.experts.3.down_proj", "model.layers.29.mlp.experts.2.down_proj", "model.layers.28.mlp.experts.3.gate_proj", "model.layers.15.mlp.experts.0.gate_proj", "model.layers.25.mlp.experts.3.up_proj", "model.layers.12.mlp.experts.2.gate_proj", "model.layers.2.mlp.experts.3.gate_proj", "model.layers.10.mlp.experts.3.down_proj", "model.layers.14.self_attn.o_proj", "model.layers.18.mlp.experts.1.up_proj", "model.layers.23.mlp.experts.3.down_proj", "model.layers.30.self_attn.o_proj", "model.layers.7.mlp.experts.2.gate_proj", "model.layers.28.self_attn.q_proj", "model.layers.5.self_attn.q_proj", "model.layers.2.mlp.experts.2.down_proj", "model.layers.4.mlp.experts.3.down_proj", "model.layers.5.self_attn.k_proj", "model.layers.15.mlp.experts.0.down_proj", "model.layers.12.mlp.experts.0.down_proj", "model.layers.7.self_attn.o_proj", "model.layers.4.mlp.experts.3.up_proj", "model.layers.8.mlp.experts.3.down_proj", "model.layers.0.mlp.experts.2.up_proj", "model.layers.1.mlp.experts.1.gate_proj", "model.layers.17.mlp.experts.2.gate_proj", "model.layers.19.mlp.experts.2.gate_proj", "model.layers.1.self_attn.q_proj", "model.layers.30.self_attn.k_proj", "model.layers.18.mlp.experts.3.up_proj", "model.layers.7.mlp.experts.0.down_proj", "model.layers.31.self_attn.k_proj", "model.layers.10.mlp.experts.2.gate_proj", "model.layers.13.self_attn.v_proj", "model.layers.16.mlp.experts.0.down_proj", "model.layers.30.mlp.experts.0.down_proj", "model.layers.23.mlp.experts.2.gate_proj", "model.layers.11.mlp.experts.0.up_proj", "model.layers.17.mlp.experts.1.up_proj", "model.layers.20.mlp.experts.3.up_proj", "model.layers.17.self_attn.k_proj", "model.layers.0.self_attn.q_proj", "model.layers.12.mlp.experts.3.gate_proj", "model.layers.1.mlp.experts.0.up_proj", "model.layers.30.self_attn.v_proj", "model.layers.5.mlp.experts.2.down_proj", "model.layers.2.mlp.experts.0.gate_proj", "model.layers.16.mlp.experts.2.up_proj", "model.layers.13.mlp.experts.3.up_proj", "model.layers.18.mlp.experts.2.down_proj", "model.layers.24.mlp.experts.2.up_proj", "model.layers.19.self_attn.o_proj", "model.layers.8.mlp.experts.1.gate_proj", "model.layers.12.self_attn.k_proj", "model.layers.20.mlp.experts.2.gate_proj", "model.layers.1.mlp.experts.1.down_proj", "model.layers.8.mlp.experts.1.down_proj", "model.layers.23.mlp.experts.0.gate_proj", "model.layers.8.mlp.experts.0.down_proj", "model.layers.2.self_attn.k_proj", "model.layers.22.mlp.experts.0.down_proj", "model.layers.1.mlp.experts.0.down_proj", "model.layers.7.mlp.experts.0.up_proj", "model.layers.0.self_attn.o_proj", "model.layers.7.self_attn.q_proj", "model.layers.10.mlp.experts.3.gate_proj", "model.layers.28.mlp.experts.3.up_proj", "model.layers.31.mlp.experts.3.up_proj", "model.layers.8.mlp.experts.1.up_proj", "model.layers.17.self_attn.o_proj", "model.layers.28.mlp.experts.1.up_proj", "model.layers.29.mlp.experts.1.up_proj", "model.layers.7.self_attn.k_proj", "model.layers.28.mlp.experts.2.gate_proj", "model.layers.14.mlp.experts.1.gate_proj", "model.layers.24.mlp.experts.1.down_proj", "model.layers.23.mlp.experts.3.up_proj", "model.layers.23.self_attn.o_proj", "model.layers.0.mlp.experts.3.up_proj", "model.layers.27.mlp.experts.1.gate_proj", "model.layers.9.mlp.experts.3.up_proj", "model.layers.10.self_attn.o_proj", "model.layers.15.self_attn.v_proj", "model.layers.30.mlp.experts.2.down_proj", "model.layers.25.mlp.experts.1.gate_proj", "model.layers.30.mlp.experts.1.gate_proj", "model.layers.2.mlp.experts.1.down_proj", "model.layers.17.mlp.experts.3.gate_proj", "model.layers.29.self_attn.k_proj", "model.layers.17.mlp.experts.3.up_proj", "model.layers.22.mlp.experts.0.gate_proj", "model.layers.7.mlp.experts.0.gate_proj", "model.layers.6.self_attn.k_proj", "model.layers.31.mlp.experts.3.gate_proj", "model.layers.13.mlp.experts.1.down_proj", "model.layers.10.self_attn.k_proj", "model.layers.16.mlp.experts.1.gate_proj", "model.layers.9.self_attn.k_proj", "model.layers.24.mlp.experts.2.down_proj", "model.layers.25.mlp.experts.0.down_proj", "model.layers.3.mlp.experts.2.gate_proj", "model.layers.12.mlp.experts.3.down_proj", "model.layers.17.self_attn.v_proj", "model.layers.22.mlp.experts.3.gate_proj", "model.layers.26.mlp.experts.2.down_proj", "model.layers.16.mlp.experts.1.down_proj", "model.layers.14.mlp.experts.1.down_proj", "model.layers.27.self_attn.v_proj", "model.layers.29.mlp.experts.2.up_proj", "model.layers.14.self_attn.k_proj", "model.layers.2.self_attn.v_proj", "model.layers.8.mlp.experts.2.down_proj", "model.layers.26.mlp.experts.3.up_proj", "model.layers.28.mlp.experts.1.gate_proj", "model.layers.3.mlp.experts.3.gate_proj", "model.layers.24.mlp.experts.2.gate_proj", "model.layers.17.mlp.experts.2.down_proj", "model.layers.20.mlp.experts.3.gate_proj", "model.layers.21.mlp.experts.0.down_proj", "model.layers.8.self_attn.k_proj", "model.layers.21.mlp.experts.2.up_proj", "model.layers.2.mlp.experts.0.up_proj", "model.layers.24.mlp.experts.3.gate_proj", "model.layers.22.mlp.experts.3.up_proj", "model.layers.3.mlp.experts.0.up_proj", "model.layers.12.mlp.experts.3.up_proj", "model.layers.20.mlp.experts.2.down_proj", "model.layers.15.mlp.experts.2.up_proj", "model.layers.25.mlp.experts.3.gate_proj", "model.layers.0.mlp.experts.3.down_proj", "model.layers.9.mlp.experts.1.down_proj", "model.layers.17.mlp.experts.2.up_proj", "model.layers.18.mlp.experts.0.down_proj", "model.layers.5.self_attn.o_proj", "model.layers.22.self_attn.k_proj", "model.layers.18.mlp.experts.0.gate_proj", "model.layers.8.mlp.experts.2.gate_proj", "model.layers.1.self_attn.v_proj", "model.layers.23.self_attn.q_proj", "model.layers.13.mlp.experts.3.gate_proj", "model.layers.26.mlp.experts.1.up_proj", "model.layers.23.mlp.experts.0.up_proj", "model.layers.12.mlp.experts.1.down_proj", "model.layers.15.self_attn.o_proj", "model.layers.2.mlp.experts.3.up_proj", "model.layers.16.mlp.experts.0.gate_proj", "model.layers.29.mlp.experts.3.gate_proj", "model.layers.15.mlp.experts.2.gate_proj", "model.layers.26.mlp.experts.2.gate_proj", "model.layers.27.mlp.experts.3.up_proj", "model.layers.10.mlp.experts.1.up_proj", "model.layers.21.self_attn.k_proj", "model.layers.3.mlp.experts.0.gate_proj", "model.layers.21.self_attn.o_proj", "model.layers.30.mlp.experts.2.up_proj", "model.layers.29.mlp.experts.1.down_proj", "model.layers.0.mlp.experts.2.gate_proj", "model.layers.16.mlp.experts.3.gate_proj", "model.layers.5.mlp.experts.3.gate_proj", "model.layers.6.mlp.experts.3.up_proj", "model.layers.2.mlp.experts.3.down_proj", "model.layers.25.self_attn.v_proj", "model.layers.12.mlp.experts.2.up_proj", "model.layers.4.mlp.experts.2.up_proj", "model.layers.0.mlp.experts.1.down_proj", "model.layers.31.self_attn.q_proj", "model.layers.6.mlp.experts.0.down_proj", "model.layers.15.mlp.experts.3.gate_proj", "model.layers.21.mlp.experts.3.gate_proj", "model.layers.25.mlp.experts.2.up_proj", "model.layers.31.self_attn.v_proj", "model.layers.14.self_attn.v_proj", "model.layers.13.mlp.experts.2.up_proj", "model.layers.15.mlp.experts.0.up_proj", "model.layers.26.mlp.experts.3.down_proj", "model.layers.31.mlp.experts.1.up_proj", "model.layers.3.mlp.experts.2.down_proj", "model.layers.20.self_attn.v_proj", "model.layers.29.mlp.experts.0.up_proj", "model.layers.11.mlp.experts.1.up_proj", "model.layers.13.mlp.experts.0.gate_proj", "model.layers.11.self_attn.q_proj", "model.layers.18.mlp.experts.0.up_proj", "model.layers.24.self_attn.v_proj", "model.layers.31.mlp.experts.0.down_proj", "model.layers.13.mlp.experts.1.up_proj", "model.layers.31.mlp.experts.0.up_proj", "model.layers.21.self_attn.q_proj", "model.layers.0.mlp.experts.0.down_proj", "model.layers.20.mlp.experts.3.down_proj", "model.layers.20.mlp.experts.0.down_proj", "model.layers.6.mlp.experts.2.gate_proj", "model.layers.3.self_attn.k_proj", "model.layers.29.mlp.experts.0.down_proj", "model.layers.2.mlp.experts.2.up_proj", "model.layers.19.self_attn.k_proj", "model.layers.1.mlp.experts.2.down_proj", "model.layers.24.self_attn.o_proj", "model.layers.23.mlp.experts.1.down_proj", "model.layers.14.mlp.experts.2.gate_proj", "model.layers.15.self_attn.q_proj", "model.layers.1.mlp.experts.0.gate_proj", "model.layers.6.mlp.experts.3.gate_proj", "model.layers.12.mlp.experts.1.gate_proj", "model.layers.1.mlp.experts.3.up_proj", "model.layers.15.mlp.experts.1.gate_proj", "model.layers.17.mlp.experts.0.up_proj", "model.layers.16.mlp.experts.0.up_proj", "model.layers.23.mlp.experts.1.gate_proj", "model.layers.24.mlp.experts.0.up_proj", "model.layers.0.mlp.experts.2.down_proj", "model.layers.7.mlp.experts.1.down_proj", "model.layers.19.mlp.experts.1.gate_proj", "model.layers.2.self_attn.o_proj", "model.layers.13.self_attn.o_proj", "model.layers.16.self_attn.v_proj", "model.layers.30.mlp.experts.1.up_proj", "model.layers.5.mlp.experts.0.gate_proj", "model.layers.5.mlp.experts.2.up_proj", "model.layers.6.mlp.experts.0.gate_proj", "model.layers.9.mlp.experts.2.down_proj", "model.layers.21.mlp.experts.1.down_proj", "model.layers.25.mlp.experts.3.down_proj", "model.layers.26.mlp.experts.2.up_proj", "model.layers.7.mlp.experts.3.down_proj", "model.layers.7.mlp.experts.1.up_proj", "model.layers.17.mlp.experts.1.down_proj", "model.layers.24.self_attn.q_proj", "model.layers.8.mlp.experts.0.up_proj", "model.layers.13.mlp.experts.2.down_proj", "model.layers.0.mlp.experts.1.gate_proj", "model.layers.31.self_attn.o_proj", "model.layers.1.mlp.experts.2.up_proj", "model.layers.11.mlp.experts.0.down_proj", "model.layers.27.mlp.experts.0.gate_proj", "model.layers.11.self_attn.k_proj", "model.layers.2.mlp.experts.0.down_proj", "model.layers.3.self_attn.v_proj", "model.layers.2.mlp.experts.1.gate_proj", "model.layers.19.mlp.experts.3.gate_proj", "model.layers.15.mlp.experts.3.down_proj", "model.layers.24.mlp.experts.1.up_proj", "model.layers.27.mlp.experts.1.down_proj", "model.layers.3.mlp.experts.2.up_proj", "model.layers.8.mlp.experts.0.gate_proj", "model.layers.31.mlp.experts.3.down_proj", "model.layers.23.mlp.experts.1.up_proj", "model.layers.9.mlp.experts.2.gate_proj", "model.layers.6.mlp.experts.2.up_proj", "model.layers.3.self_attn.o_proj", "model.layers.31.mlp.experts.1.gate_proj" ], "task_type": "CAUSAL_LM" }