{ "metadata": { "total_size": 10107595008 }, "weight_map": { "lm_head.weight": "model-00003-of-00003.safetensors", "model.embed_tokens.weight": "model-00001-of-00003.safetensors", "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.0.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.0.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.0.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.1.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.1.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.1.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.1.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.1.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.1.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.10.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.10.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.10.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.10.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.10.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.10.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.11.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.11.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.11.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.11.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.11.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.11.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.12.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.12.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.12.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.12.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.12.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.12.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.13.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.13.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.13.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.13.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.13.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.13.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.13.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.14.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.14.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.14.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.14.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.14.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.14.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.15.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.15.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.15.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.15.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.15.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.15.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.16.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.16.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.16.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.16.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.16.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.16.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.17.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.17.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.17.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.17.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.17.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.17.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.18.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.18.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.18.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.18.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.18.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.18.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.19.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.19.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.19.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.19.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.19.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.19.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.2.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.2.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.2.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.2.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.2.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.2.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.20.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.20.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.20.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.20.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.20.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.20.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.21.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.21.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.21.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.21.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.21.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.21.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.22.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.22.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.22.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.22.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.22.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.22.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.23.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.23.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.23.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.23.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.23.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.23.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.24.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.24.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.24.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.24.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.24.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.24.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.25.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.25.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.25.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.25.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.25.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.25.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.0.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.1.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.10.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.11.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.12.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.13.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.14.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.15.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.2.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.3.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.4.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.5.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.6.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.7.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.8.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.down_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.down_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.down_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.down_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.gate_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.gate_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.gate_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.gate_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.mask_up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.up_proj.fused_layer.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.up_proj.qa_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.up_proj.qb_weights": "model-00002-of-00003.safetensors", "model.layers.26.mlp.experts.9.up_proj.scaling_factor": "model-00002-of-00003.safetensors", "model.layers.26.mlp.gate.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.inv_mapping_dict": "model-00002-of-00003.safetensors", "model.layers.26.mlp.shared_experts.down_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.shared_experts.gate_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.mlp.shared_experts.up_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.26.self_attn.kv_a_layernorm.weight": "model-00002-of-00003.safetensors", "model.layers.26.self_attn.kv_a_proj_with_mqa.weight": "model-00002-of-00003.safetensors", "model.layers.26.self_attn.kv_b_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.3.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.3.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.3.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.3.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.3.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.3.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.4.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.4.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.4.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.4.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.4.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.4.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.5.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.5.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.5.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.5.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.5.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.5.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.6.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.6.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.6.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.6.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.6.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.6.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.7.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.7.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.7.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.7.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.7.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.7.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.8.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.8.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.8.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.8.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.8.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.8.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.0.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.1.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.10.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.11.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.12.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.13.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.14.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.15.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.2.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.3.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.4.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.5.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.6.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.7.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.8.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.down_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.down_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.down_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.down_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.gate_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.gate_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.gate_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.gate_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.mask_up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.up_proj.fused_layer.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.up_proj.qa_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.up_proj.qb_weights": "model-00001-of-00003.safetensors", "model.layers.9.mlp.experts.9.up_proj.scaling_factor": "model-00001-of-00003.safetensors", "model.layers.9.mlp.gate.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.inv_mapping_dict": "model-00001-of-00003.safetensors", "model.layers.9.mlp.shared_experts.down_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.shared_experts.gate_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.mlp.shared_experts.up_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.9.self_attn.kv_a_layernorm.weight": "model-00001-of-00003.safetensors", "model.layers.9.self_attn.kv_a_proj_with_mqa.weight": "model-00001-of-00003.safetensors", "model.layers.9.self_attn.kv_b_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", "model.norm.weight": "model-00002-of-00003.safetensors" } }