| { | |
| "architectures": [ | |
| "Ernie4_5_MoeForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_ernie4_5_moe.Ernie4_5_MoeConfig", | |
| "AutoModel": "modeling_ernie4_5_moe.Ernie4_5_Model", | |
| "AutoModelForCausalLM": "modeling_ernie4_5_moe.Ernie4_5_MoeForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 2560, | |
| "intermediate_size": 12288, | |
| "max_position_embeddings": 131072, | |
| "model_type": "ernie4_5_moe", | |
| "moe_capacity": [ | |
| 64, | |
| 64, | |
| 64 | |
| ], | |
| "moe_intermediate_size": 1536, | |
| "moe_k": 6, | |
| "moe_layer_interval": 1, | |
| "moe_layer_start_index": 1, | |
| "moe_num_experts": 64, | |
| "moe_num_shared_experts": 2, | |
| "moe_use_aux_free": true, | |
| "multi_token_pred_lambda": 1.0, | |
| "num_attention_heads": 20, | |
| "num_hidden_layers": 28, | |
| "num_key_value_heads": 4, | |
| "num_nextn_predict_layers": 1, | |
| "pad_token_id": 0, | |
| "quantization": { | |
| "group_size": 64, | |
| "bits": 5, | |
| "model.embed_tokens": true, | |
| "model.layers.0.self_attn.q_proj": true, | |
| "model.layers.0.self_attn.k_proj": true, | |
| "model.layers.0.self_attn.v_proj": true, | |
| "model.layers.0.self_attn.o_proj": true, | |
| "model.layers.0.self_attn.rope": false, | |
| "model.layers.0.mlp.gate_proj": true, | |
| "model.layers.0.mlp.down_proj": true, | |
| "model.layers.0.mlp.up_proj": true, | |
| "model.layers.0.input_layernorm": false, | |
| "model.layers.0.post_attention_layernorm": false, | |
| "model.layers.1.self_attn.q_proj": true, | |
| "model.layers.1.self_attn.k_proj": true, | |
| "model.layers.1.self_attn.v_proj": true, | |
| "model.layers.1.self_attn.o_proj": true, | |
| "model.layers.1.self_attn.rope": false, | |
| "model.layers.1.mlp.gate": true, | |
| "model.layers.1.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.1.mlp.switch_mlp.up_proj": true, | |
| "model.layers.1.mlp.switch_mlp.down_proj": true, | |
| "model.layers.1.mlp.switch_mlp.activation": false, | |
| "model.layers.1.mlp.shared_experts.gate_proj": true, | |
| "model.layers.1.mlp.shared_experts.down_proj": true, | |
| "model.layers.1.mlp.shared_experts.up_proj": true, | |
| "model.layers.1.mlp.moe_statics": false, | |
| "model.layers.1.mlp.gate_act": false, | |
| "model.layers.1.input_layernorm": false, | |
| "model.layers.1.post_attention_layernorm": false, | |
| "model.layers.2.self_attn.q_proj": true, | |
| "model.layers.2.self_attn.k_proj": true, | |
| "model.layers.2.self_attn.v_proj": true, | |
| "model.layers.2.self_attn.o_proj": true, | |
| "model.layers.2.self_attn.rope": false, | |
| "model.layers.2.mlp.gate": true, | |
| "model.layers.2.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.2.mlp.switch_mlp.up_proj": true, | |
| "model.layers.2.mlp.switch_mlp.down_proj": true, | |
| "model.layers.2.mlp.switch_mlp.activation": false, | |
| "model.layers.2.mlp.shared_experts.gate_proj": true, | |
| "model.layers.2.mlp.shared_experts.down_proj": true, | |
| "model.layers.2.mlp.shared_experts.up_proj": true, | |
| "model.layers.2.mlp.moe_statics": false, | |
| "model.layers.2.mlp.gate_act": false, | |
| "model.layers.2.input_layernorm": false, | |
| "model.layers.2.post_attention_layernorm": false, | |
| "model.layers.3.self_attn.q_proj": true, | |
| "model.layers.3.self_attn.k_proj": true, | |
| "model.layers.3.self_attn.v_proj": true, | |
| "model.layers.3.self_attn.o_proj": true, | |
| "model.layers.3.self_attn.rope": false, | |
| "model.layers.3.mlp.gate": true, | |
| "model.layers.3.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.3.mlp.switch_mlp.up_proj": true, | |
| "model.layers.3.mlp.switch_mlp.down_proj": true, | |
| "model.layers.3.mlp.switch_mlp.activation": false, | |
| "model.layers.3.mlp.shared_experts.gate_proj": true, | |
| "model.layers.3.mlp.shared_experts.down_proj": true, | |
| "model.layers.3.mlp.shared_experts.up_proj": true, | |
| "model.layers.3.mlp.moe_statics": false, | |
| "model.layers.3.mlp.gate_act": false, | |
| "model.layers.3.input_layernorm": false, | |
| "model.layers.3.post_attention_layernorm": false, | |
| "model.layers.4.self_attn.q_proj": true, | |
| "model.layers.4.self_attn.k_proj": true, | |
| "model.layers.4.self_attn.v_proj": true, | |
| "model.layers.4.self_attn.o_proj": true, | |
| "model.layers.4.self_attn.rope": false, | |
| "model.layers.4.mlp.gate": true, | |
| "model.layers.4.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.4.mlp.switch_mlp.up_proj": true, | |
| "model.layers.4.mlp.switch_mlp.down_proj": true, | |
| "model.layers.4.mlp.switch_mlp.activation": false, | |
| "model.layers.4.mlp.shared_experts.gate_proj": true, | |
| "model.layers.4.mlp.shared_experts.down_proj": true, | |
| "model.layers.4.mlp.shared_experts.up_proj": true, | |
| "model.layers.4.mlp.moe_statics": false, | |
| "model.layers.4.mlp.gate_act": false, | |
| "model.layers.4.input_layernorm": false, | |
| "model.layers.4.post_attention_layernorm": false, | |
| "model.layers.5.self_attn.q_proj": true, | |
| "model.layers.5.self_attn.k_proj": true, | |
| "model.layers.5.self_attn.v_proj": true, | |
| "model.layers.5.self_attn.o_proj": true, | |
| "model.layers.5.self_attn.rope": false, | |
| "model.layers.5.mlp.gate": true, | |
| "model.layers.5.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.5.mlp.switch_mlp.up_proj": true, | |
| "model.layers.5.mlp.switch_mlp.down_proj": true, | |
| "model.layers.5.mlp.switch_mlp.activation": false, | |
| "model.layers.5.mlp.shared_experts.gate_proj": true, | |
| "model.layers.5.mlp.shared_experts.down_proj": true, | |
| "model.layers.5.mlp.shared_experts.up_proj": true, | |
| "model.layers.5.mlp.moe_statics": false, | |
| "model.layers.5.mlp.gate_act": false, | |
| "model.layers.5.input_layernorm": false, | |
| "model.layers.5.post_attention_layernorm": false, | |
| "model.layers.6.self_attn.q_proj": true, | |
| "model.layers.6.self_attn.k_proj": true, | |
| "model.layers.6.self_attn.v_proj": true, | |
| "model.layers.6.self_attn.o_proj": true, | |
| "model.layers.6.self_attn.rope": false, | |
| "model.layers.6.mlp.gate": true, | |
| "model.layers.6.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.6.mlp.switch_mlp.up_proj": true, | |
| "model.layers.6.mlp.switch_mlp.down_proj": true, | |
| "model.layers.6.mlp.switch_mlp.activation": false, | |
| "model.layers.6.mlp.shared_experts.gate_proj": true, | |
| "model.layers.6.mlp.shared_experts.down_proj": true, | |
| "model.layers.6.mlp.shared_experts.up_proj": true, | |
| "model.layers.6.mlp.moe_statics": false, | |
| "model.layers.6.mlp.gate_act": false, | |
| "model.layers.6.input_layernorm": false, | |
| "model.layers.6.post_attention_layernorm": false, | |
| "model.layers.7.self_attn.q_proj": true, | |
| "model.layers.7.self_attn.k_proj": true, | |
| "model.layers.7.self_attn.v_proj": true, | |
| "model.layers.7.self_attn.o_proj": true, | |
| "model.layers.7.self_attn.rope": false, | |
| "model.layers.7.mlp.gate": true, | |
| "model.layers.7.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.7.mlp.switch_mlp.up_proj": true, | |
| "model.layers.7.mlp.switch_mlp.down_proj": true, | |
| "model.layers.7.mlp.switch_mlp.activation": false, | |
| "model.layers.7.mlp.shared_experts.gate_proj": true, | |
| "model.layers.7.mlp.shared_experts.down_proj": true, | |
| "model.layers.7.mlp.shared_experts.up_proj": true, | |
| "model.layers.7.mlp.moe_statics": false, | |
| "model.layers.7.mlp.gate_act": false, | |
| "model.layers.7.input_layernorm": false, | |
| "model.layers.7.post_attention_layernorm": false, | |
| "model.layers.8.self_attn.q_proj": true, | |
| "model.layers.8.self_attn.k_proj": true, | |
| "model.layers.8.self_attn.v_proj": true, | |
| "model.layers.8.self_attn.o_proj": true, | |
| "model.layers.8.self_attn.rope": false, | |
| "model.layers.8.mlp.gate": true, | |
| "model.layers.8.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.8.mlp.switch_mlp.up_proj": true, | |
| "model.layers.8.mlp.switch_mlp.down_proj": true, | |
| "model.layers.8.mlp.switch_mlp.activation": false, | |
| "model.layers.8.mlp.shared_experts.gate_proj": true, | |
| "model.layers.8.mlp.shared_experts.down_proj": true, | |
| "model.layers.8.mlp.shared_experts.up_proj": true, | |
| "model.layers.8.mlp.moe_statics": false, | |
| "model.layers.8.mlp.gate_act": false, | |
| "model.layers.8.input_layernorm": false, | |
| "model.layers.8.post_attention_layernorm": false, | |
| "model.layers.9.self_attn.q_proj": true, | |
| "model.layers.9.self_attn.k_proj": true, | |
| "model.layers.9.self_attn.v_proj": true, | |
| "model.layers.9.self_attn.o_proj": true, | |
| "model.layers.9.self_attn.rope": false, | |
| "model.layers.9.mlp.gate": true, | |
| "model.layers.9.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.9.mlp.switch_mlp.up_proj": true, | |
| "model.layers.9.mlp.switch_mlp.down_proj": true, | |
| "model.layers.9.mlp.switch_mlp.activation": false, | |
| "model.layers.9.mlp.shared_experts.gate_proj": true, | |
| "model.layers.9.mlp.shared_experts.down_proj": true, | |
| "model.layers.9.mlp.shared_experts.up_proj": true, | |
| "model.layers.9.mlp.moe_statics": false, | |
| "model.layers.9.mlp.gate_act": false, | |
| "model.layers.9.input_layernorm": false, | |
| "model.layers.9.post_attention_layernorm": false, | |
| "model.layers.10.self_attn.q_proj": true, | |
| "model.layers.10.self_attn.k_proj": true, | |
| "model.layers.10.self_attn.v_proj": true, | |
| "model.layers.10.self_attn.o_proj": true, | |
| "model.layers.10.self_attn.rope": false, | |
| "model.layers.10.mlp.gate": true, | |
| "model.layers.10.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.10.mlp.switch_mlp.up_proj": true, | |
| "model.layers.10.mlp.switch_mlp.down_proj": true, | |
| "model.layers.10.mlp.switch_mlp.activation": false, | |
| "model.layers.10.mlp.shared_experts.gate_proj": true, | |
| "model.layers.10.mlp.shared_experts.down_proj": true, | |
| "model.layers.10.mlp.shared_experts.up_proj": true, | |
| "model.layers.10.mlp.moe_statics": false, | |
| "model.layers.10.mlp.gate_act": false, | |
| "model.layers.10.input_layernorm": false, | |
| "model.layers.10.post_attention_layernorm": false, | |
| "model.layers.11.self_attn.q_proj": true, | |
| "model.layers.11.self_attn.k_proj": true, | |
| "model.layers.11.self_attn.v_proj": true, | |
| "model.layers.11.self_attn.o_proj": true, | |
| "model.layers.11.self_attn.rope": false, | |
| "model.layers.11.mlp.gate": true, | |
| "model.layers.11.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.11.mlp.switch_mlp.up_proj": true, | |
| "model.layers.11.mlp.switch_mlp.down_proj": true, | |
| "model.layers.11.mlp.switch_mlp.activation": false, | |
| "model.layers.11.mlp.shared_experts.gate_proj": true, | |
| "model.layers.11.mlp.shared_experts.down_proj": true, | |
| "model.layers.11.mlp.shared_experts.up_proj": true, | |
| "model.layers.11.mlp.moe_statics": false, | |
| "model.layers.11.mlp.gate_act": false, | |
| "model.layers.11.input_layernorm": false, | |
| "model.layers.11.post_attention_layernorm": false, | |
| "model.layers.12.self_attn.q_proj": true, | |
| "model.layers.12.self_attn.k_proj": true, | |
| "model.layers.12.self_attn.v_proj": true, | |
| "model.layers.12.self_attn.o_proj": true, | |
| "model.layers.12.self_attn.rope": false, | |
| "model.layers.12.mlp.gate": true, | |
| "model.layers.12.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.12.mlp.switch_mlp.up_proj": true, | |
| "model.layers.12.mlp.switch_mlp.down_proj": true, | |
| "model.layers.12.mlp.switch_mlp.activation": false, | |
| "model.layers.12.mlp.shared_experts.gate_proj": true, | |
| "model.layers.12.mlp.shared_experts.down_proj": true, | |
| "model.layers.12.mlp.shared_experts.up_proj": true, | |
| "model.layers.12.mlp.moe_statics": false, | |
| "model.layers.12.mlp.gate_act": false, | |
| "model.layers.12.input_layernorm": false, | |
| "model.layers.12.post_attention_layernorm": false, | |
| "model.layers.13.self_attn.q_proj": true, | |
| "model.layers.13.self_attn.k_proj": true, | |
| "model.layers.13.self_attn.v_proj": true, | |
| "model.layers.13.self_attn.o_proj": true, | |
| "model.layers.13.self_attn.rope": false, | |
| "model.layers.13.mlp.gate": true, | |
| "model.layers.13.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.13.mlp.switch_mlp.up_proj": true, | |
| "model.layers.13.mlp.switch_mlp.down_proj": true, | |
| "model.layers.13.mlp.switch_mlp.activation": false, | |
| "model.layers.13.mlp.shared_experts.gate_proj": true, | |
| "model.layers.13.mlp.shared_experts.down_proj": true, | |
| "model.layers.13.mlp.shared_experts.up_proj": true, | |
| "model.layers.13.mlp.moe_statics": false, | |
| "model.layers.13.mlp.gate_act": false, | |
| "model.layers.13.input_layernorm": false, | |
| "model.layers.13.post_attention_layernorm": false, | |
| "model.layers.14.self_attn.q_proj": true, | |
| "model.layers.14.self_attn.k_proj": true, | |
| "model.layers.14.self_attn.v_proj": true, | |
| "model.layers.14.self_attn.o_proj": true, | |
| "model.layers.14.self_attn.rope": false, | |
| "model.layers.14.mlp.gate": true, | |
| "model.layers.14.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.14.mlp.switch_mlp.up_proj": true, | |
| "model.layers.14.mlp.switch_mlp.down_proj": true, | |
| "model.layers.14.mlp.switch_mlp.activation": false, | |
| "model.layers.14.mlp.shared_experts.gate_proj": true, | |
| "model.layers.14.mlp.shared_experts.down_proj": true, | |
| "model.layers.14.mlp.shared_experts.up_proj": true, | |
| "model.layers.14.mlp.moe_statics": false, | |
| "model.layers.14.mlp.gate_act": false, | |
| "model.layers.14.input_layernorm": false, | |
| "model.layers.14.post_attention_layernorm": false, | |
| "model.layers.15.self_attn.q_proj": true, | |
| "model.layers.15.self_attn.k_proj": true, | |
| "model.layers.15.self_attn.v_proj": true, | |
| "model.layers.15.self_attn.o_proj": true, | |
| "model.layers.15.self_attn.rope": false, | |
| "model.layers.15.mlp.gate": true, | |
| "model.layers.15.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.15.mlp.switch_mlp.up_proj": true, | |
| "model.layers.15.mlp.switch_mlp.down_proj": true, | |
| "model.layers.15.mlp.switch_mlp.activation": false, | |
| "model.layers.15.mlp.shared_experts.gate_proj": true, | |
| "model.layers.15.mlp.shared_experts.down_proj": true, | |
| "model.layers.15.mlp.shared_experts.up_proj": true, | |
| "model.layers.15.mlp.moe_statics": false, | |
| "model.layers.15.mlp.gate_act": false, | |
| "model.layers.15.input_layernorm": false, | |
| "model.layers.15.post_attention_layernorm": false, | |
| "model.layers.16.self_attn.q_proj": true, | |
| "model.layers.16.self_attn.k_proj": true, | |
| "model.layers.16.self_attn.v_proj": true, | |
| "model.layers.16.self_attn.o_proj": true, | |
| "model.layers.16.self_attn.rope": false, | |
| "model.layers.16.mlp.gate": true, | |
| "model.layers.16.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.16.mlp.switch_mlp.up_proj": true, | |
| "model.layers.16.mlp.switch_mlp.down_proj": true, | |
| "model.layers.16.mlp.switch_mlp.activation": false, | |
| "model.layers.16.mlp.shared_experts.gate_proj": true, | |
| "model.layers.16.mlp.shared_experts.down_proj": true, | |
| "model.layers.16.mlp.shared_experts.up_proj": true, | |
| "model.layers.16.mlp.moe_statics": false, | |
| "model.layers.16.mlp.gate_act": false, | |
| "model.layers.16.input_layernorm": false, | |
| "model.layers.16.post_attention_layernorm": false, | |
| "model.layers.17.self_attn.q_proj": true, | |
| "model.layers.17.self_attn.k_proj": true, | |
| "model.layers.17.self_attn.v_proj": true, | |
| "model.layers.17.self_attn.o_proj": true, | |
| "model.layers.17.self_attn.rope": false, | |
| "model.layers.17.mlp.gate": true, | |
| "model.layers.17.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.17.mlp.switch_mlp.up_proj": true, | |
| "model.layers.17.mlp.switch_mlp.down_proj": true, | |
| "model.layers.17.mlp.switch_mlp.activation": false, | |
| "model.layers.17.mlp.shared_experts.gate_proj": true, | |
| "model.layers.17.mlp.shared_experts.down_proj": true, | |
| "model.layers.17.mlp.shared_experts.up_proj": true, | |
| "model.layers.17.mlp.moe_statics": false, | |
| "model.layers.17.mlp.gate_act": false, | |
| "model.layers.17.input_layernorm": false, | |
| "model.layers.17.post_attention_layernorm": false, | |
| "model.layers.18.self_attn.q_proj": true, | |
| "model.layers.18.self_attn.k_proj": true, | |
| "model.layers.18.self_attn.v_proj": true, | |
| "model.layers.18.self_attn.o_proj": true, | |
| "model.layers.18.self_attn.rope": false, | |
| "model.layers.18.mlp.gate": true, | |
| "model.layers.18.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.18.mlp.switch_mlp.up_proj": true, | |
| "model.layers.18.mlp.switch_mlp.down_proj": true, | |
| "model.layers.18.mlp.switch_mlp.activation": false, | |
| "model.layers.18.mlp.shared_experts.gate_proj": true, | |
| "model.layers.18.mlp.shared_experts.down_proj": true, | |
| "model.layers.18.mlp.shared_experts.up_proj": true, | |
| "model.layers.18.mlp.moe_statics": false, | |
| "model.layers.18.mlp.gate_act": false, | |
| "model.layers.18.input_layernorm": false, | |
| "model.layers.18.post_attention_layernorm": false, | |
| "model.layers.19.self_attn.q_proj": true, | |
| "model.layers.19.self_attn.k_proj": true, | |
| "model.layers.19.self_attn.v_proj": true, | |
| "model.layers.19.self_attn.o_proj": true, | |
| "model.layers.19.self_attn.rope": false, | |
| "model.layers.19.mlp.gate": true, | |
| "model.layers.19.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.19.mlp.switch_mlp.up_proj": true, | |
| "model.layers.19.mlp.switch_mlp.down_proj": true, | |
| "model.layers.19.mlp.switch_mlp.activation": false, | |
| "model.layers.19.mlp.shared_experts.gate_proj": true, | |
| "model.layers.19.mlp.shared_experts.down_proj": true, | |
| "model.layers.19.mlp.shared_experts.up_proj": true, | |
| "model.layers.19.mlp.moe_statics": false, | |
| "model.layers.19.mlp.gate_act": false, | |
| "model.layers.19.input_layernorm": false, | |
| "model.layers.19.post_attention_layernorm": false, | |
| "model.layers.20.self_attn.q_proj": true, | |
| "model.layers.20.self_attn.k_proj": true, | |
| "model.layers.20.self_attn.v_proj": true, | |
| "model.layers.20.self_attn.o_proj": true, | |
| "model.layers.20.self_attn.rope": false, | |
| "model.layers.20.mlp.gate": true, | |
| "model.layers.20.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.20.mlp.switch_mlp.up_proj": true, | |
| "model.layers.20.mlp.switch_mlp.down_proj": true, | |
| "model.layers.20.mlp.switch_mlp.activation": false, | |
| "model.layers.20.mlp.shared_experts.gate_proj": true, | |
| "model.layers.20.mlp.shared_experts.down_proj": true, | |
| "model.layers.20.mlp.shared_experts.up_proj": true, | |
| "model.layers.20.mlp.moe_statics": false, | |
| "model.layers.20.mlp.gate_act": false, | |
| "model.layers.20.input_layernorm": false, | |
| "model.layers.20.post_attention_layernorm": false, | |
| "model.layers.21.self_attn.q_proj": true, | |
| "model.layers.21.self_attn.k_proj": true, | |
| "model.layers.21.self_attn.v_proj": true, | |
| "model.layers.21.self_attn.o_proj": true, | |
| "model.layers.21.self_attn.rope": false, | |
| "model.layers.21.mlp.gate": true, | |
| "model.layers.21.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.21.mlp.switch_mlp.up_proj": true, | |
| "model.layers.21.mlp.switch_mlp.down_proj": true, | |
| "model.layers.21.mlp.switch_mlp.activation": false, | |
| "model.layers.21.mlp.shared_experts.gate_proj": true, | |
| "model.layers.21.mlp.shared_experts.down_proj": true, | |
| "model.layers.21.mlp.shared_experts.up_proj": true, | |
| "model.layers.21.mlp.moe_statics": false, | |
| "model.layers.21.mlp.gate_act": false, | |
| "model.layers.21.input_layernorm": false, | |
| "model.layers.21.post_attention_layernorm": false, | |
| "model.layers.22.self_attn.q_proj": true, | |
| "model.layers.22.self_attn.k_proj": true, | |
| "model.layers.22.self_attn.v_proj": true, | |
| "model.layers.22.self_attn.o_proj": true, | |
| "model.layers.22.self_attn.rope": false, | |
| "model.layers.22.mlp.gate": true, | |
| "model.layers.22.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.22.mlp.switch_mlp.up_proj": true, | |
| "model.layers.22.mlp.switch_mlp.down_proj": true, | |
| "model.layers.22.mlp.switch_mlp.activation": false, | |
| "model.layers.22.mlp.shared_experts.gate_proj": true, | |
| "model.layers.22.mlp.shared_experts.down_proj": true, | |
| "model.layers.22.mlp.shared_experts.up_proj": true, | |
| "model.layers.22.mlp.moe_statics": false, | |
| "model.layers.22.mlp.gate_act": false, | |
| "model.layers.22.input_layernorm": false, | |
| "model.layers.22.post_attention_layernorm": false, | |
| "model.layers.23.self_attn.q_proj": true, | |
| "model.layers.23.self_attn.k_proj": true, | |
| "model.layers.23.self_attn.v_proj": true, | |
| "model.layers.23.self_attn.o_proj": true, | |
| "model.layers.23.self_attn.rope": false, | |
| "model.layers.23.mlp.gate": true, | |
| "model.layers.23.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.23.mlp.switch_mlp.up_proj": true, | |
| "model.layers.23.mlp.switch_mlp.down_proj": true, | |
| "model.layers.23.mlp.switch_mlp.activation": false, | |
| "model.layers.23.mlp.shared_experts.gate_proj": true, | |
| "model.layers.23.mlp.shared_experts.down_proj": true, | |
| "model.layers.23.mlp.shared_experts.up_proj": true, | |
| "model.layers.23.mlp.moe_statics": false, | |
| "model.layers.23.mlp.gate_act": false, | |
| "model.layers.23.input_layernorm": false, | |
| "model.layers.23.post_attention_layernorm": false, | |
| "model.layers.24.self_attn.q_proj": true, | |
| "model.layers.24.self_attn.k_proj": true, | |
| "model.layers.24.self_attn.v_proj": true, | |
| "model.layers.24.self_attn.o_proj": true, | |
| "model.layers.24.self_attn.rope": false, | |
| "model.layers.24.mlp.gate": true, | |
| "model.layers.24.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.24.mlp.switch_mlp.up_proj": true, | |
| "model.layers.24.mlp.switch_mlp.down_proj": true, | |
| "model.layers.24.mlp.switch_mlp.activation": false, | |
| "model.layers.24.mlp.shared_experts.gate_proj": true, | |
| "model.layers.24.mlp.shared_experts.down_proj": true, | |
| "model.layers.24.mlp.shared_experts.up_proj": true, | |
| "model.layers.24.mlp.moe_statics": false, | |
| "model.layers.24.mlp.gate_act": false, | |
| "model.layers.24.input_layernorm": false, | |
| "model.layers.24.post_attention_layernorm": false, | |
| "model.layers.25.self_attn.q_proj": true, | |
| "model.layers.25.self_attn.k_proj": true, | |
| "model.layers.25.self_attn.v_proj": true, | |
| "model.layers.25.self_attn.o_proj": true, | |
| "model.layers.25.self_attn.rope": false, | |
| "model.layers.25.mlp.gate": true, | |
| "model.layers.25.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.25.mlp.switch_mlp.up_proj": true, | |
| "model.layers.25.mlp.switch_mlp.down_proj": true, | |
| "model.layers.25.mlp.switch_mlp.activation": false, | |
| "model.layers.25.mlp.shared_experts.gate_proj": true, | |
| "model.layers.25.mlp.shared_experts.down_proj": true, | |
| "model.layers.25.mlp.shared_experts.up_proj": true, | |
| "model.layers.25.mlp.moe_statics": false, | |
| "model.layers.25.mlp.gate_act": false, | |
| "model.layers.25.input_layernorm": false, | |
| "model.layers.25.post_attention_layernorm": false, | |
| "model.layers.26.self_attn.q_proj": true, | |
| "model.layers.26.self_attn.k_proj": true, | |
| "model.layers.26.self_attn.v_proj": true, | |
| "model.layers.26.self_attn.o_proj": true, | |
| "model.layers.26.self_attn.rope": false, | |
| "model.layers.26.mlp.gate": true, | |
| "model.layers.26.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.26.mlp.switch_mlp.up_proj": true, | |
| "model.layers.26.mlp.switch_mlp.down_proj": true, | |
| "model.layers.26.mlp.switch_mlp.activation": false, | |
| "model.layers.26.mlp.shared_experts.gate_proj": true, | |
| "model.layers.26.mlp.shared_experts.down_proj": true, | |
| "model.layers.26.mlp.shared_experts.up_proj": true, | |
| "model.layers.26.mlp.moe_statics": false, | |
| "model.layers.26.mlp.gate_act": false, | |
| "model.layers.26.input_layernorm": false, | |
| "model.layers.26.post_attention_layernorm": false, | |
| "model.layers.27.self_attn.q_proj": true, | |
| "model.layers.27.self_attn.k_proj": true, | |
| "model.layers.27.self_attn.v_proj": true, | |
| "model.layers.27.self_attn.o_proj": true, | |
| "model.layers.27.self_attn.rope": false, | |
| "model.layers.27.mlp.gate": true, | |
| "model.layers.27.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.27.mlp.switch_mlp.up_proj": true, | |
| "model.layers.27.mlp.switch_mlp.down_proj": true, | |
| "model.layers.27.mlp.switch_mlp.activation": false, | |
| "model.layers.27.mlp.shared_experts.gate_proj": true, | |
| "model.layers.27.mlp.shared_experts.down_proj": true, | |
| "model.layers.27.mlp.shared_experts.up_proj": true, | |
| "model.layers.27.mlp.moe_statics": false, | |
| "model.layers.27.mlp.gate_act": false, | |
| "model.layers.27.input_layernorm": false, | |
| "model.layers.27.post_attention_layernorm": false, | |
| "model.norm": false | |
| }, | |
| "quantization_config": { | |
| "group_size": 64, | |
| "bits": 5, | |
| "model.embed_tokens": true, | |
| "model.layers.0.self_attn.q_proj": true, | |
| "model.layers.0.self_attn.k_proj": true, | |
| "model.layers.0.self_attn.v_proj": true, | |
| "model.layers.0.self_attn.o_proj": true, | |
| "model.layers.0.self_attn.rope": false, | |
| "model.layers.0.mlp.gate_proj": true, | |
| "model.layers.0.mlp.down_proj": true, | |
| "model.layers.0.mlp.up_proj": true, | |
| "model.layers.0.input_layernorm": false, | |
| "model.layers.0.post_attention_layernorm": false, | |
| "model.layers.1.self_attn.q_proj": true, | |
| "model.layers.1.self_attn.k_proj": true, | |
| "model.layers.1.self_attn.v_proj": true, | |
| "model.layers.1.self_attn.o_proj": true, | |
| "model.layers.1.self_attn.rope": false, | |
| "model.layers.1.mlp.gate": true, | |
| "model.layers.1.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.1.mlp.switch_mlp.up_proj": true, | |
| "model.layers.1.mlp.switch_mlp.down_proj": true, | |
| "model.layers.1.mlp.switch_mlp.activation": false, | |
| "model.layers.1.mlp.shared_experts.gate_proj": true, | |
| "model.layers.1.mlp.shared_experts.down_proj": true, | |
| "model.layers.1.mlp.shared_experts.up_proj": true, | |
| "model.layers.1.mlp.moe_statics": false, | |
| "model.layers.1.mlp.gate_act": false, | |
| "model.layers.1.input_layernorm": false, | |
| "model.layers.1.post_attention_layernorm": false, | |
| "model.layers.2.self_attn.q_proj": true, | |
| "model.layers.2.self_attn.k_proj": true, | |
| "model.layers.2.self_attn.v_proj": true, | |
| "model.layers.2.self_attn.o_proj": true, | |
| "model.layers.2.self_attn.rope": false, | |
| "model.layers.2.mlp.gate": true, | |
| "model.layers.2.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.2.mlp.switch_mlp.up_proj": true, | |
| "model.layers.2.mlp.switch_mlp.down_proj": true, | |
| "model.layers.2.mlp.switch_mlp.activation": false, | |
| "model.layers.2.mlp.shared_experts.gate_proj": true, | |
| "model.layers.2.mlp.shared_experts.down_proj": true, | |
| "model.layers.2.mlp.shared_experts.up_proj": true, | |
| "model.layers.2.mlp.moe_statics": false, | |
| "model.layers.2.mlp.gate_act": false, | |
| "model.layers.2.input_layernorm": false, | |
| "model.layers.2.post_attention_layernorm": false, | |
| "model.layers.3.self_attn.q_proj": true, | |
| "model.layers.3.self_attn.k_proj": true, | |
| "model.layers.3.self_attn.v_proj": true, | |
| "model.layers.3.self_attn.o_proj": true, | |
| "model.layers.3.self_attn.rope": false, | |
| "model.layers.3.mlp.gate": true, | |
| "model.layers.3.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.3.mlp.switch_mlp.up_proj": true, | |
| "model.layers.3.mlp.switch_mlp.down_proj": true, | |
| "model.layers.3.mlp.switch_mlp.activation": false, | |
| "model.layers.3.mlp.shared_experts.gate_proj": true, | |
| "model.layers.3.mlp.shared_experts.down_proj": true, | |
| "model.layers.3.mlp.shared_experts.up_proj": true, | |
| "model.layers.3.mlp.moe_statics": false, | |
| "model.layers.3.mlp.gate_act": false, | |
| "model.layers.3.input_layernorm": false, | |
| "model.layers.3.post_attention_layernorm": false, | |
| "model.layers.4.self_attn.q_proj": true, | |
| "model.layers.4.self_attn.k_proj": true, | |
| "model.layers.4.self_attn.v_proj": true, | |
| "model.layers.4.self_attn.o_proj": true, | |
| "model.layers.4.self_attn.rope": false, | |
| "model.layers.4.mlp.gate": true, | |
| "model.layers.4.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.4.mlp.switch_mlp.up_proj": true, | |
| "model.layers.4.mlp.switch_mlp.down_proj": true, | |
| "model.layers.4.mlp.switch_mlp.activation": false, | |
| "model.layers.4.mlp.shared_experts.gate_proj": true, | |
| "model.layers.4.mlp.shared_experts.down_proj": true, | |
| "model.layers.4.mlp.shared_experts.up_proj": true, | |
| "model.layers.4.mlp.moe_statics": false, | |
| "model.layers.4.mlp.gate_act": false, | |
| "model.layers.4.input_layernorm": false, | |
| "model.layers.4.post_attention_layernorm": false, | |
| "model.layers.5.self_attn.q_proj": true, | |
| "model.layers.5.self_attn.k_proj": true, | |
| "model.layers.5.self_attn.v_proj": true, | |
| "model.layers.5.self_attn.o_proj": true, | |
| "model.layers.5.self_attn.rope": false, | |
| "model.layers.5.mlp.gate": true, | |
| "model.layers.5.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.5.mlp.switch_mlp.up_proj": true, | |
| "model.layers.5.mlp.switch_mlp.down_proj": true, | |
| "model.layers.5.mlp.switch_mlp.activation": false, | |
| "model.layers.5.mlp.shared_experts.gate_proj": true, | |
| "model.layers.5.mlp.shared_experts.down_proj": true, | |
| "model.layers.5.mlp.shared_experts.up_proj": true, | |
| "model.layers.5.mlp.moe_statics": false, | |
| "model.layers.5.mlp.gate_act": false, | |
| "model.layers.5.input_layernorm": false, | |
| "model.layers.5.post_attention_layernorm": false, | |
| "model.layers.6.self_attn.q_proj": true, | |
| "model.layers.6.self_attn.k_proj": true, | |
| "model.layers.6.self_attn.v_proj": true, | |
| "model.layers.6.self_attn.o_proj": true, | |
| "model.layers.6.self_attn.rope": false, | |
| "model.layers.6.mlp.gate": true, | |
| "model.layers.6.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.6.mlp.switch_mlp.up_proj": true, | |
| "model.layers.6.mlp.switch_mlp.down_proj": true, | |
| "model.layers.6.mlp.switch_mlp.activation": false, | |
| "model.layers.6.mlp.shared_experts.gate_proj": true, | |
| "model.layers.6.mlp.shared_experts.down_proj": true, | |
| "model.layers.6.mlp.shared_experts.up_proj": true, | |
| "model.layers.6.mlp.moe_statics": false, | |
| "model.layers.6.mlp.gate_act": false, | |
| "model.layers.6.input_layernorm": false, | |
| "model.layers.6.post_attention_layernorm": false, | |
| "model.layers.7.self_attn.q_proj": true, | |
| "model.layers.7.self_attn.k_proj": true, | |
| "model.layers.7.self_attn.v_proj": true, | |
| "model.layers.7.self_attn.o_proj": true, | |
| "model.layers.7.self_attn.rope": false, | |
| "model.layers.7.mlp.gate": true, | |
| "model.layers.7.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.7.mlp.switch_mlp.up_proj": true, | |
| "model.layers.7.mlp.switch_mlp.down_proj": true, | |
| "model.layers.7.mlp.switch_mlp.activation": false, | |
| "model.layers.7.mlp.shared_experts.gate_proj": true, | |
| "model.layers.7.mlp.shared_experts.down_proj": true, | |
| "model.layers.7.mlp.shared_experts.up_proj": true, | |
| "model.layers.7.mlp.moe_statics": false, | |
| "model.layers.7.mlp.gate_act": false, | |
| "model.layers.7.input_layernorm": false, | |
| "model.layers.7.post_attention_layernorm": false, | |
| "model.layers.8.self_attn.q_proj": true, | |
| "model.layers.8.self_attn.k_proj": true, | |
| "model.layers.8.self_attn.v_proj": true, | |
| "model.layers.8.self_attn.o_proj": true, | |
| "model.layers.8.self_attn.rope": false, | |
| "model.layers.8.mlp.gate": true, | |
| "model.layers.8.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.8.mlp.switch_mlp.up_proj": true, | |
| "model.layers.8.mlp.switch_mlp.down_proj": true, | |
| "model.layers.8.mlp.switch_mlp.activation": false, | |
| "model.layers.8.mlp.shared_experts.gate_proj": true, | |
| "model.layers.8.mlp.shared_experts.down_proj": true, | |
| "model.layers.8.mlp.shared_experts.up_proj": true, | |
| "model.layers.8.mlp.moe_statics": false, | |
| "model.layers.8.mlp.gate_act": false, | |
| "model.layers.8.input_layernorm": false, | |
| "model.layers.8.post_attention_layernorm": false, | |
| "model.layers.9.self_attn.q_proj": true, | |
| "model.layers.9.self_attn.k_proj": true, | |
| "model.layers.9.self_attn.v_proj": true, | |
| "model.layers.9.self_attn.o_proj": true, | |
| "model.layers.9.self_attn.rope": false, | |
| "model.layers.9.mlp.gate": true, | |
| "model.layers.9.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.9.mlp.switch_mlp.up_proj": true, | |
| "model.layers.9.mlp.switch_mlp.down_proj": true, | |
| "model.layers.9.mlp.switch_mlp.activation": false, | |
| "model.layers.9.mlp.shared_experts.gate_proj": true, | |
| "model.layers.9.mlp.shared_experts.down_proj": true, | |
| "model.layers.9.mlp.shared_experts.up_proj": true, | |
| "model.layers.9.mlp.moe_statics": false, | |
| "model.layers.9.mlp.gate_act": false, | |
| "model.layers.9.input_layernorm": false, | |
| "model.layers.9.post_attention_layernorm": false, | |
| "model.layers.10.self_attn.q_proj": true, | |
| "model.layers.10.self_attn.k_proj": true, | |
| "model.layers.10.self_attn.v_proj": true, | |
| "model.layers.10.self_attn.o_proj": true, | |
| "model.layers.10.self_attn.rope": false, | |
| "model.layers.10.mlp.gate": true, | |
| "model.layers.10.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.10.mlp.switch_mlp.up_proj": true, | |
| "model.layers.10.mlp.switch_mlp.down_proj": true, | |
| "model.layers.10.mlp.switch_mlp.activation": false, | |
| "model.layers.10.mlp.shared_experts.gate_proj": true, | |
| "model.layers.10.mlp.shared_experts.down_proj": true, | |
| "model.layers.10.mlp.shared_experts.up_proj": true, | |
| "model.layers.10.mlp.moe_statics": false, | |
| "model.layers.10.mlp.gate_act": false, | |
| "model.layers.10.input_layernorm": false, | |
| "model.layers.10.post_attention_layernorm": false, | |
| "model.layers.11.self_attn.q_proj": true, | |
| "model.layers.11.self_attn.k_proj": true, | |
| "model.layers.11.self_attn.v_proj": true, | |
| "model.layers.11.self_attn.o_proj": true, | |
| "model.layers.11.self_attn.rope": false, | |
| "model.layers.11.mlp.gate": true, | |
| "model.layers.11.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.11.mlp.switch_mlp.up_proj": true, | |
| "model.layers.11.mlp.switch_mlp.down_proj": true, | |
| "model.layers.11.mlp.switch_mlp.activation": false, | |
| "model.layers.11.mlp.shared_experts.gate_proj": true, | |
| "model.layers.11.mlp.shared_experts.down_proj": true, | |
| "model.layers.11.mlp.shared_experts.up_proj": true, | |
| "model.layers.11.mlp.moe_statics": false, | |
| "model.layers.11.mlp.gate_act": false, | |
| "model.layers.11.input_layernorm": false, | |
| "model.layers.11.post_attention_layernorm": false, | |
| "model.layers.12.self_attn.q_proj": true, | |
| "model.layers.12.self_attn.k_proj": true, | |
| "model.layers.12.self_attn.v_proj": true, | |
| "model.layers.12.self_attn.o_proj": true, | |
| "model.layers.12.self_attn.rope": false, | |
| "model.layers.12.mlp.gate": true, | |
| "model.layers.12.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.12.mlp.switch_mlp.up_proj": true, | |
| "model.layers.12.mlp.switch_mlp.down_proj": true, | |
| "model.layers.12.mlp.switch_mlp.activation": false, | |
| "model.layers.12.mlp.shared_experts.gate_proj": true, | |
| "model.layers.12.mlp.shared_experts.down_proj": true, | |
| "model.layers.12.mlp.shared_experts.up_proj": true, | |
| "model.layers.12.mlp.moe_statics": false, | |
| "model.layers.12.mlp.gate_act": false, | |
| "model.layers.12.input_layernorm": false, | |
| "model.layers.12.post_attention_layernorm": false, | |
| "model.layers.13.self_attn.q_proj": true, | |
| "model.layers.13.self_attn.k_proj": true, | |
| "model.layers.13.self_attn.v_proj": true, | |
| "model.layers.13.self_attn.o_proj": true, | |
| "model.layers.13.self_attn.rope": false, | |
| "model.layers.13.mlp.gate": true, | |
| "model.layers.13.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.13.mlp.switch_mlp.up_proj": true, | |
| "model.layers.13.mlp.switch_mlp.down_proj": true, | |
| "model.layers.13.mlp.switch_mlp.activation": false, | |
| "model.layers.13.mlp.shared_experts.gate_proj": true, | |
| "model.layers.13.mlp.shared_experts.down_proj": true, | |
| "model.layers.13.mlp.shared_experts.up_proj": true, | |
| "model.layers.13.mlp.moe_statics": false, | |
| "model.layers.13.mlp.gate_act": false, | |
| "model.layers.13.input_layernorm": false, | |
| "model.layers.13.post_attention_layernorm": false, | |
| "model.layers.14.self_attn.q_proj": true, | |
| "model.layers.14.self_attn.k_proj": true, | |
| "model.layers.14.self_attn.v_proj": true, | |
| "model.layers.14.self_attn.o_proj": true, | |
| "model.layers.14.self_attn.rope": false, | |
| "model.layers.14.mlp.gate": true, | |
| "model.layers.14.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.14.mlp.switch_mlp.up_proj": true, | |
| "model.layers.14.mlp.switch_mlp.down_proj": true, | |
| "model.layers.14.mlp.switch_mlp.activation": false, | |
| "model.layers.14.mlp.shared_experts.gate_proj": true, | |
| "model.layers.14.mlp.shared_experts.down_proj": true, | |
| "model.layers.14.mlp.shared_experts.up_proj": true, | |
| "model.layers.14.mlp.moe_statics": false, | |
| "model.layers.14.mlp.gate_act": false, | |
| "model.layers.14.input_layernorm": false, | |
| "model.layers.14.post_attention_layernorm": false, | |
| "model.layers.15.self_attn.q_proj": true, | |
| "model.layers.15.self_attn.k_proj": true, | |
| "model.layers.15.self_attn.v_proj": true, | |
| "model.layers.15.self_attn.o_proj": true, | |
| "model.layers.15.self_attn.rope": false, | |
| "model.layers.15.mlp.gate": true, | |
| "model.layers.15.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.15.mlp.switch_mlp.up_proj": true, | |
| "model.layers.15.mlp.switch_mlp.down_proj": true, | |
| "model.layers.15.mlp.switch_mlp.activation": false, | |
| "model.layers.15.mlp.shared_experts.gate_proj": true, | |
| "model.layers.15.mlp.shared_experts.down_proj": true, | |
| "model.layers.15.mlp.shared_experts.up_proj": true, | |
| "model.layers.15.mlp.moe_statics": false, | |
| "model.layers.15.mlp.gate_act": false, | |
| "model.layers.15.input_layernorm": false, | |
| "model.layers.15.post_attention_layernorm": false, | |
| "model.layers.16.self_attn.q_proj": true, | |
| "model.layers.16.self_attn.k_proj": true, | |
| "model.layers.16.self_attn.v_proj": true, | |
| "model.layers.16.self_attn.o_proj": true, | |
| "model.layers.16.self_attn.rope": false, | |
| "model.layers.16.mlp.gate": true, | |
| "model.layers.16.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.16.mlp.switch_mlp.up_proj": true, | |
| "model.layers.16.mlp.switch_mlp.down_proj": true, | |
| "model.layers.16.mlp.switch_mlp.activation": false, | |
| "model.layers.16.mlp.shared_experts.gate_proj": true, | |
| "model.layers.16.mlp.shared_experts.down_proj": true, | |
| "model.layers.16.mlp.shared_experts.up_proj": true, | |
| "model.layers.16.mlp.moe_statics": false, | |
| "model.layers.16.mlp.gate_act": false, | |
| "model.layers.16.input_layernorm": false, | |
| "model.layers.16.post_attention_layernorm": false, | |
| "model.layers.17.self_attn.q_proj": true, | |
| "model.layers.17.self_attn.k_proj": true, | |
| "model.layers.17.self_attn.v_proj": true, | |
| "model.layers.17.self_attn.o_proj": true, | |
| "model.layers.17.self_attn.rope": false, | |
| "model.layers.17.mlp.gate": true, | |
| "model.layers.17.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.17.mlp.switch_mlp.up_proj": true, | |
| "model.layers.17.mlp.switch_mlp.down_proj": true, | |
| "model.layers.17.mlp.switch_mlp.activation": false, | |
| "model.layers.17.mlp.shared_experts.gate_proj": true, | |
| "model.layers.17.mlp.shared_experts.down_proj": true, | |
| "model.layers.17.mlp.shared_experts.up_proj": true, | |
| "model.layers.17.mlp.moe_statics": false, | |
| "model.layers.17.mlp.gate_act": false, | |
| "model.layers.17.input_layernorm": false, | |
| "model.layers.17.post_attention_layernorm": false, | |
| "model.layers.18.self_attn.q_proj": true, | |
| "model.layers.18.self_attn.k_proj": true, | |
| "model.layers.18.self_attn.v_proj": true, | |
| "model.layers.18.self_attn.o_proj": true, | |
| "model.layers.18.self_attn.rope": false, | |
| "model.layers.18.mlp.gate": true, | |
| "model.layers.18.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.18.mlp.switch_mlp.up_proj": true, | |
| "model.layers.18.mlp.switch_mlp.down_proj": true, | |
| "model.layers.18.mlp.switch_mlp.activation": false, | |
| "model.layers.18.mlp.shared_experts.gate_proj": true, | |
| "model.layers.18.mlp.shared_experts.down_proj": true, | |
| "model.layers.18.mlp.shared_experts.up_proj": true, | |
| "model.layers.18.mlp.moe_statics": false, | |
| "model.layers.18.mlp.gate_act": false, | |
| "model.layers.18.input_layernorm": false, | |
| "model.layers.18.post_attention_layernorm": false, | |
| "model.layers.19.self_attn.q_proj": true, | |
| "model.layers.19.self_attn.k_proj": true, | |
| "model.layers.19.self_attn.v_proj": true, | |
| "model.layers.19.self_attn.o_proj": true, | |
| "model.layers.19.self_attn.rope": false, | |
| "model.layers.19.mlp.gate": true, | |
| "model.layers.19.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.19.mlp.switch_mlp.up_proj": true, | |
| "model.layers.19.mlp.switch_mlp.down_proj": true, | |
| "model.layers.19.mlp.switch_mlp.activation": false, | |
| "model.layers.19.mlp.shared_experts.gate_proj": true, | |
| "model.layers.19.mlp.shared_experts.down_proj": true, | |
| "model.layers.19.mlp.shared_experts.up_proj": true, | |
| "model.layers.19.mlp.moe_statics": false, | |
| "model.layers.19.mlp.gate_act": false, | |
| "model.layers.19.input_layernorm": false, | |
| "model.layers.19.post_attention_layernorm": false, | |
| "model.layers.20.self_attn.q_proj": true, | |
| "model.layers.20.self_attn.k_proj": true, | |
| "model.layers.20.self_attn.v_proj": true, | |
| "model.layers.20.self_attn.o_proj": true, | |
| "model.layers.20.self_attn.rope": false, | |
| "model.layers.20.mlp.gate": true, | |
| "model.layers.20.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.20.mlp.switch_mlp.up_proj": true, | |
| "model.layers.20.mlp.switch_mlp.down_proj": true, | |
| "model.layers.20.mlp.switch_mlp.activation": false, | |
| "model.layers.20.mlp.shared_experts.gate_proj": true, | |
| "model.layers.20.mlp.shared_experts.down_proj": true, | |
| "model.layers.20.mlp.shared_experts.up_proj": true, | |
| "model.layers.20.mlp.moe_statics": false, | |
| "model.layers.20.mlp.gate_act": false, | |
| "model.layers.20.input_layernorm": false, | |
| "model.layers.20.post_attention_layernorm": false, | |
| "model.layers.21.self_attn.q_proj": true, | |
| "model.layers.21.self_attn.k_proj": true, | |
| "model.layers.21.self_attn.v_proj": true, | |
| "model.layers.21.self_attn.o_proj": true, | |
| "model.layers.21.self_attn.rope": false, | |
| "model.layers.21.mlp.gate": true, | |
| "model.layers.21.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.21.mlp.switch_mlp.up_proj": true, | |
| "model.layers.21.mlp.switch_mlp.down_proj": true, | |
| "model.layers.21.mlp.switch_mlp.activation": false, | |
| "model.layers.21.mlp.shared_experts.gate_proj": true, | |
| "model.layers.21.mlp.shared_experts.down_proj": true, | |
| "model.layers.21.mlp.shared_experts.up_proj": true, | |
| "model.layers.21.mlp.moe_statics": false, | |
| "model.layers.21.mlp.gate_act": false, | |
| "model.layers.21.input_layernorm": false, | |
| "model.layers.21.post_attention_layernorm": false, | |
| "model.layers.22.self_attn.q_proj": true, | |
| "model.layers.22.self_attn.k_proj": true, | |
| "model.layers.22.self_attn.v_proj": true, | |
| "model.layers.22.self_attn.o_proj": true, | |
| "model.layers.22.self_attn.rope": false, | |
| "model.layers.22.mlp.gate": true, | |
| "model.layers.22.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.22.mlp.switch_mlp.up_proj": true, | |
| "model.layers.22.mlp.switch_mlp.down_proj": true, | |
| "model.layers.22.mlp.switch_mlp.activation": false, | |
| "model.layers.22.mlp.shared_experts.gate_proj": true, | |
| "model.layers.22.mlp.shared_experts.down_proj": true, | |
| "model.layers.22.mlp.shared_experts.up_proj": true, | |
| "model.layers.22.mlp.moe_statics": false, | |
| "model.layers.22.mlp.gate_act": false, | |
| "model.layers.22.input_layernorm": false, | |
| "model.layers.22.post_attention_layernorm": false, | |
| "model.layers.23.self_attn.q_proj": true, | |
| "model.layers.23.self_attn.k_proj": true, | |
| "model.layers.23.self_attn.v_proj": true, | |
| "model.layers.23.self_attn.o_proj": true, | |
| "model.layers.23.self_attn.rope": false, | |
| "model.layers.23.mlp.gate": true, | |
| "model.layers.23.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.23.mlp.switch_mlp.up_proj": true, | |
| "model.layers.23.mlp.switch_mlp.down_proj": true, | |
| "model.layers.23.mlp.switch_mlp.activation": false, | |
| "model.layers.23.mlp.shared_experts.gate_proj": true, | |
| "model.layers.23.mlp.shared_experts.down_proj": true, | |
| "model.layers.23.mlp.shared_experts.up_proj": true, | |
| "model.layers.23.mlp.moe_statics": false, | |
| "model.layers.23.mlp.gate_act": false, | |
| "model.layers.23.input_layernorm": false, | |
| "model.layers.23.post_attention_layernorm": false, | |
| "model.layers.24.self_attn.q_proj": true, | |
| "model.layers.24.self_attn.k_proj": true, | |
| "model.layers.24.self_attn.v_proj": true, | |
| "model.layers.24.self_attn.o_proj": true, | |
| "model.layers.24.self_attn.rope": false, | |
| "model.layers.24.mlp.gate": true, | |
| "model.layers.24.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.24.mlp.switch_mlp.up_proj": true, | |
| "model.layers.24.mlp.switch_mlp.down_proj": true, | |
| "model.layers.24.mlp.switch_mlp.activation": false, | |
| "model.layers.24.mlp.shared_experts.gate_proj": true, | |
| "model.layers.24.mlp.shared_experts.down_proj": true, | |
| "model.layers.24.mlp.shared_experts.up_proj": true, | |
| "model.layers.24.mlp.moe_statics": false, | |
| "model.layers.24.mlp.gate_act": false, | |
| "model.layers.24.input_layernorm": false, | |
| "model.layers.24.post_attention_layernorm": false, | |
| "model.layers.25.self_attn.q_proj": true, | |
| "model.layers.25.self_attn.k_proj": true, | |
| "model.layers.25.self_attn.v_proj": true, | |
| "model.layers.25.self_attn.o_proj": true, | |
| "model.layers.25.self_attn.rope": false, | |
| "model.layers.25.mlp.gate": true, | |
| "model.layers.25.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.25.mlp.switch_mlp.up_proj": true, | |
| "model.layers.25.mlp.switch_mlp.down_proj": true, | |
| "model.layers.25.mlp.switch_mlp.activation": false, | |
| "model.layers.25.mlp.shared_experts.gate_proj": true, | |
| "model.layers.25.mlp.shared_experts.down_proj": true, | |
| "model.layers.25.mlp.shared_experts.up_proj": true, | |
| "model.layers.25.mlp.moe_statics": false, | |
| "model.layers.25.mlp.gate_act": false, | |
| "model.layers.25.input_layernorm": false, | |
| "model.layers.25.post_attention_layernorm": false, | |
| "model.layers.26.self_attn.q_proj": true, | |
| "model.layers.26.self_attn.k_proj": true, | |
| "model.layers.26.self_attn.v_proj": true, | |
| "model.layers.26.self_attn.o_proj": true, | |
| "model.layers.26.self_attn.rope": false, | |
| "model.layers.26.mlp.gate": true, | |
| "model.layers.26.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.26.mlp.switch_mlp.up_proj": true, | |
| "model.layers.26.mlp.switch_mlp.down_proj": true, | |
| "model.layers.26.mlp.switch_mlp.activation": false, | |
| "model.layers.26.mlp.shared_experts.gate_proj": true, | |
| "model.layers.26.mlp.shared_experts.down_proj": true, | |
| "model.layers.26.mlp.shared_experts.up_proj": true, | |
| "model.layers.26.mlp.moe_statics": false, | |
| "model.layers.26.mlp.gate_act": false, | |
| "model.layers.26.input_layernorm": false, | |
| "model.layers.26.post_attention_layernorm": false, | |
| "model.layers.27.self_attn.q_proj": true, | |
| "model.layers.27.self_attn.k_proj": true, | |
| "model.layers.27.self_attn.v_proj": true, | |
| "model.layers.27.self_attn.o_proj": true, | |
| "model.layers.27.self_attn.rope": false, | |
| "model.layers.27.mlp.gate": true, | |
| "model.layers.27.mlp.switch_mlp.gate_proj": true, | |
| "model.layers.27.mlp.switch_mlp.up_proj": true, | |
| "model.layers.27.mlp.switch_mlp.down_proj": true, | |
| "model.layers.27.mlp.switch_mlp.activation": false, | |
| "model.layers.27.mlp.shared_experts.gate_proj": true, | |
| "model.layers.27.mlp.shared_experts.down_proj": true, | |
| "model.layers.27.mlp.shared_experts.up_proj": true, | |
| "model.layers.27.mlp.moe_statics": false, | |
| "model.layers.27.mlp.gate_act": false, | |
| "model.layers.27.input_layernorm": false, | |
| "model.layers.27.post_attention_layernorm": false, | |
| "model.norm": false | |
| }, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 500000, | |
| "tie_word_embeddings": true, | |
| "torch_dtype": "bfloat16", | |
| "use_bias": false, | |
| "use_cache": false, | |
| "vocab_size": 103424 | |
| } |