pytorch_model.bin.index.json

{"metadata": {"total_size": 15354717664}, "weight_map": {"model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.gate.wg.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.16.constant": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.17.constant": "pytorch_model-00001-of-00025.bin", "model.layers.0.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00001-of-00025.bin", "model.layers.1.input_layernorm.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.gate.wg.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.gate.gate_map.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.16.constant": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.17.constant": "pytorch_model-00002-of-00025.bin", "model.layers.1.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00002-of-00025.bin", "model.layers.2.input_layernorm.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.gate.wg.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.gate.gate_map.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.16.constant": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.17.constant": "pytorch_model-00003-of-00025.bin", "model.layers.2.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00003-of-00025.bin", "model.layers.3.input_layernorm.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.gate.wg.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.gate.gate_map.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.16.constant": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.17.constant": "pytorch_model-00004-of-00025.bin", "model.layers.3.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00004-of-00025.bin", "model.layers.4.input_layernorm.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.gate.wg.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.gate.gate_map.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.16.constant": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.17.constant": "pytorch_model-00005-of-00025.bin", "model.layers.4.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00005-of-00025.bin", "model.layers.5.input_layernorm.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.gate.wg.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.gate.gate_map.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.16.constant": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.17.constant": "pytorch_model-00006-of-00025.bin", "model.layers.5.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00006-of-00025.bin", "model.layers.6.input_layernorm.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.gate.wg.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.gate.gate_map.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.16.constant": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.17.constant": "pytorch_model-00007-of-00025.bin", "model.layers.6.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00007-of-00025.bin", "model.layers.7.input_layernorm.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.gate.wg.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.gate.gate_map.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.16.constant": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.17.constant": "pytorch_model-00008-of-00025.bin", "model.layers.7.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00008-of-00025.bin", "model.layers.8.input_layernorm.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.gate.wg.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.gate.gate_map.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.16.constant": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.17.constant": "pytorch_model-00009-of-00025.bin", "model.layers.8.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00009-of-00025.bin", "model.layers.9.input_layernorm.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.gate.wg.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.gate.gate_map.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.16.constant": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.17.constant": "pytorch_model-00010-of-00025.bin", "model.layers.9.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00010-of-00025.bin", "model.layers.10.input_layernorm.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.gate.wg.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.gate.gate_map.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.16.constant": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.17.constant": "pytorch_model-00011-of-00025.bin", "model.layers.10.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00011-of-00025.bin", "model.layers.11.input_layernorm.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.gate.wg.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.gate.gate_map.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.16.constant": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.17.constant": "pytorch_model-00012-of-00025.bin", "model.layers.11.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00012-of-00025.bin", "model.layers.12.input_layernorm.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.gate.wg.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.gate.gate_map.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.16.constant": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.17.constant": "pytorch_model-00013-of-00025.bin", "model.layers.12.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00013-of-00025.bin", "model.layers.13.input_layernorm.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.gate.wg.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.gate.gate_map.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.16.constant": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.17.constant": "pytorch_model-00014-of-00025.bin", "model.layers.13.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00014-of-00025.bin", "model.layers.14.input_layernorm.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.self_attn.rotary_emb.inv_freq": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.gate.wg.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.gate.gate_map.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.16.constant": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.17.constant": "pytorch_model-00015-of-00025.bin", "model.layers.14.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00015-of-00025.bin", "model.layers.15.input_layernorm.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.self_attn.rotary_emb.inv_freq": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.gate.wg.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.gate.gate_map.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.16.constant": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.17.constant": "pytorch_model-00016-of-00025.bin", "model.layers.15.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00016-of-00025.bin", "model.layers.16.input_layernorm.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.self_attn.rotary_emb.inv_freq": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.gate.wg.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.gate.gate_map.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.16.constant": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.17.constant": "pytorch_model-00017-of-00025.bin", "model.layers.16.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00017-of-00025.bin", "model.layers.17.input_layernorm.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.self_attn.rotary_emb.inv_freq": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.gate.wg.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.gate.gate_map.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.16.constant": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.17.constant": "pytorch_model-00018-of-00025.bin", "model.layers.17.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00018-of-00025.bin", "model.layers.18.input_layernorm.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.self_attn.rotary_emb.inv_freq": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.gate.wg.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.gate.gate_map.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.16.constant": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.17.constant": "pytorch_model-00019-of-00025.bin", "model.layers.18.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00019-of-00025.bin", "model.layers.19.input_layernorm.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.self_attn.rotary_emb.inv_freq": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.gate.wg.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.gate.gate_map.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.16.constant": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.17.constant": "pytorch_model-00020-of-00025.bin", "model.layers.19.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00020-of-00025.bin", "model.layers.20.input_layernorm.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.self_attn.rotary_emb.inv_freq": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.gate.wg.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.gate.gate_map.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.16.constant": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.17.constant": "pytorch_model-00021-of-00025.bin", "model.layers.20.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00021-of-00025.bin", "model.layers.21.input_layernorm.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.self_attn.rotary_emb.inv_freq": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.gate.wg.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.gate.gate_map.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.16.constant": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.17.constant": "pytorch_model-00022-of-00025.bin", "model.layers.21.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00022-of-00025.bin", "model.layers.22.input_layernorm.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.self_attn.rotary_emb.inv_freq": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.gate.wg.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.gate.gate_map.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.16.constant": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.17.constant": "pytorch_model-00023-of-00025.bin", "model.layers.22.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00023-of-00025.bin", "model.layers.23.input_layernorm.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.self_attn.rotary_emb.inv_freq": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.gate.wg.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.gate.gate_map.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.0.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.0.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.1.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.1.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.2.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.2.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.3.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.3.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.4.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.4.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.5.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.5.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.6.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.6.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.7.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.7.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.8.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.8.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.9.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.9.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.10.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.10.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.11.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.11.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.12.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.12.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.13.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.13.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.14.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.14.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.15.up_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.15.down_proj.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.16.constant": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.16.wg.weight": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.17.constant": "pytorch_model-00024-of-00025.bin", "model.layers.23.mlp.moe.experts.experts.17.wg.weight": "pytorch_model-00024-of-00025.bin", "model.norm.weight": "pytorch_model-00025-of-00025.bin", "model.embed_tokens.weight": "pytorch_model-00025-of-00025.bin", "lm_head.weight": "pytorch_model-00025-of-00025.bin"}}