{
  "metadata": {
    "total_parameters": 34561885184,
    "total_size": 69123856384
  },
  "weight_map": {
    "lm_head.out_proj.weight": "model-00015-of-00015.safetensors",
    "model.decoder.layers.0.cross_attn.k_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.cross_attn.o_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.cross_attn.q_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.cross_attn.v_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.experts.bias": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.experts.mlp.v1": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.experts.mlp.w1": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.experts.mlp.w2": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.router.layer.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.mlp.router.top_k_bias": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.post_cross_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.post_self_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.pre_self_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.self_attn.k_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.self_attn.o_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.self_attn.q_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.0.self_attn.v_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.cross_attn.k_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.cross_attn.o_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.cross_attn.q_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.cross_attn.v_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.mlp.experts.bias": "model-00010-of-00015.safetensors",
    "model.decoder.layers.1.mlp.experts.mlp.v1": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.mlp.experts.mlp.w1": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.mlp.experts.mlp.w2": "model-00010-of-00015.safetensors",
    "model.decoder.layers.1.mlp.router.layer.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.1.mlp.router.top_k_bias": "model-00010-of-00015.safetensors",
    "model.decoder.layers.1.post_cross_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.post_self_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.pre_self_attn_layernorm.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.self_attn.k_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.self_attn.o_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.self_attn.q_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.1.self_attn.v_proj.weight": "model-00009-of-00015.safetensors",
    "model.decoder.layers.10.cross_attn.k_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.cross_attn.o_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.cross_attn.q_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.cross_attn.v_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.experts.bias": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.experts.mlp.v1": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.experts.mlp.w1": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.experts.mlp.w2": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.router.layer.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.mlp.router.top_k_bias": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.post_cross_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.post_self_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.pre_self_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.self_attn.k_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.self_attn.o_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.self_attn.q_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.10.self_attn.v_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.cross_attn.k_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.cross_attn.o_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.cross_attn.q_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.cross_attn.v_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.mlp.experts.bias": "model-00015-of-00015.safetensors",
    "model.decoder.layers.11.mlp.experts.mlp.v1": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.mlp.experts.mlp.w1": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.mlp.experts.mlp.w2": "model-00015-of-00015.safetensors",
    "model.decoder.layers.11.mlp.router.layer.weight": "model-00015-of-00015.safetensors",
    "model.decoder.layers.11.mlp.router.top_k_bias": "model-00015-of-00015.safetensors",
    "model.decoder.layers.11.post_cross_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.post_self_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.pre_self_attn_layernorm.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.self_attn.k_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.self_attn.o_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.self_attn.q_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.11.self_attn.v_proj.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.2.cross_attn.k_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.cross_attn.o_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.cross_attn.q_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.cross_attn.v_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.experts.bias": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.experts.mlp.v1": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.experts.mlp.w1": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.experts.mlp.w2": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.router.layer.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.mlp.router.top_k_bias": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.post_cross_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.post_self_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.pre_self_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.self_attn.k_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.self_attn.o_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.self_attn.q_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.2.self_attn.v_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.cross_attn.k_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.cross_attn.o_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.cross_attn.q_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.cross_attn.v_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.mlp.experts.bias": "model-00011-of-00015.safetensors",
    "model.decoder.layers.3.mlp.experts.mlp.v1": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.mlp.experts.mlp.w1": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.mlp.experts.mlp.w2": "model-00011-of-00015.safetensors",
    "model.decoder.layers.3.mlp.router.layer.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.3.mlp.router.top_k_bias": "model-00011-of-00015.safetensors",
    "model.decoder.layers.3.post_cross_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.post_self_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.pre_self_attn_layernorm.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.self_attn.k_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.self_attn.o_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.self_attn.q_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.3.self_attn.v_proj.weight": "model-00010-of-00015.safetensors",
    "model.decoder.layers.4.cross_attn.k_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.cross_attn.o_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.cross_attn.q_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.cross_attn.v_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.experts.bias": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.experts.mlp.v1": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.experts.mlp.w1": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.experts.mlp.w2": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.router.layer.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.mlp.router.top_k_bias": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.post_cross_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.post_self_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.pre_self_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.self_attn.k_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.self_attn.o_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.self_attn.q_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.4.self_attn.v_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.cross_attn.k_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.cross_attn.o_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.cross_attn.q_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.cross_attn.v_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.mlp.experts.bias": "model-00012-of-00015.safetensors",
    "model.decoder.layers.5.mlp.experts.mlp.v1": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.mlp.experts.mlp.w1": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.mlp.experts.mlp.w2": "model-00012-of-00015.safetensors",
    "model.decoder.layers.5.mlp.router.layer.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.5.mlp.router.top_k_bias": "model-00012-of-00015.safetensors",
    "model.decoder.layers.5.post_cross_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.post_self_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.pre_self_attn_layernorm.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.self_attn.k_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.self_attn.o_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.self_attn.q_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.5.self_attn.v_proj.weight": "model-00011-of-00015.safetensors",
    "model.decoder.layers.6.cross_attn.k_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.cross_attn.o_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.cross_attn.q_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.cross_attn.v_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.experts.bias": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.experts.mlp.v1": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.experts.mlp.w1": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.experts.mlp.w2": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.router.layer.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.mlp.router.top_k_bias": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.post_cross_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.post_self_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.pre_self_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.self_attn.k_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.self_attn.o_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.self_attn.q_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.6.self_attn.v_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.cross_attn.k_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.cross_attn.o_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.cross_attn.q_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.cross_attn.v_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.mlp.experts.bias": "model-00013-of-00015.safetensors",
    "model.decoder.layers.7.mlp.experts.mlp.v1": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.mlp.experts.mlp.w1": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.mlp.experts.mlp.w2": "model-00013-of-00015.safetensors",
    "model.decoder.layers.7.mlp.router.layer.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.7.mlp.router.top_k_bias": "model-00013-of-00015.safetensors",
    "model.decoder.layers.7.post_cross_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.post_self_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.pre_self_attn_layernorm.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.self_attn.k_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.self_attn.o_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.self_attn.q_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.7.self_attn.v_proj.weight": "model-00012-of-00015.safetensors",
    "model.decoder.layers.8.cross_attn.k_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.cross_attn.o_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.cross_attn.q_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.cross_attn.v_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.experts.bias": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.experts.mlp.v1": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.experts.mlp.w1": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.experts.mlp.w2": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.router.layer.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.mlp.router.top_k_bias": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.post_cross_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.post_self_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.pre_self_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.self_attn.k_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.self_attn.o_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.self_attn.q_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.8.self_attn.v_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.cross_attn.k_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.cross_attn.o_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.cross_attn.q_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.cross_attn.v_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.mlp.experts.bias": "model-00014-of-00015.safetensors",
    "model.decoder.layers.9.mlp.experts.mlp.v1": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.mlp.experts.mlp.w1": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.mlp.experts.mlp.w2": "model-00014-of-00015.safetensors",
    "model.decoder.layers.9.mlp.router.layer.weight": "model-00014-of-00015.safetensors",
    "model.decoder.layers.9.mlp.router.top_k_bias": "model-00014-of-00015.safetensors",
    "model.decoder.layers.9.post_cross_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.post_self_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.pre_self_attn_layernorm.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.self_attn.k_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.self_attn.o_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.self_attn.q_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.layers.9.self_attn.v_proj.weight": "model-00013-of-00015.safetensors",
    "model.decoder.norm.weight": "model-00015-of-00015.safetensors",
    "model.encoder.embed_tokens.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.experts.bias": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.experts.mlp.v1": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.experts.mlp.w1": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.experts.mlp.w2": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.router.layer.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.mlp.router.top_k_bias": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.post_self_attn_layernorm.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.pre_self_attn_layernorm.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.self_attn.o_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.mlp.experts.bias": "model-00002-of-00015.safetensors",
    "model.encoder.layers.1.mlp.experts.mlp.v1": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.mlp.experts.mlp.w1": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.mlp.experts.mlp.w2": "model-00002-of-00015.safetensors",
    "model.encoder.layers.1.mlp.router.layer.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.1.mlp.router.top_k_bias": "model-00002-of-00015.safetensors",
    "model.encoder.layers.1.post_self_attn_layernorm.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.pre_self_attn_layernorm.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.self_attn.o_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00015.safetensors",
    "model.encoder.layers.10.mlp.experts.bias": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.mlp.experts.mlp.v1": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.mlp.experts.mlp.w1": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.mlp.experts.mlp.w2": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.mlp.router.layer.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.mlp.router.top_k_bias": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.post_self_attn_layernorm.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.pre_self_attn_layernorm.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.self_attn.k_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.self_attn.o_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.self_attn.q_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.10.self_attn.v_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.mlp.experts.bias": "model-00007-of-00015.safetensors",
    "model.encoder.layers.11.mlp.experts.mlp.v1": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.mlp.experts.mlp.w1": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.mlp.experts.mlp.w2": "model-00007-of-00015.safetensors",
    "model.encoder.layers.11.mlp.router.layer.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.11.mlp.router.top_k_bias": "model-00007-of-00015.safetensors",
    "model.encoder.layers.11.post_self_attn_layernorm.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.pre_self_attn_layernorm.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.self_attn.k_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.self_attn.o_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.self_attn.q_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.11.self_attn.v_proj.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.12.mlp.experts.bias": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.mlp.experts.mlp.v1": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.mlp.experts.mlp.w1": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.mlp.experts.mlp.w2": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.mlp.router.layer.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.mlp.router.top_k_bias": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.post_self_attn_layernorm.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.pre_self_attn_layernorm.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.self_attn.k_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.self_attn.o_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.self_attn.q_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.12.self_attn.v_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.mlp.experts.bias": "model-00008-of-00015.safetensors",
    "model.encoder.layers.13.mlp.experts.mlp.v1": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.mlp.experts.mlp.w1": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.mlp.experts.mlp.w2": "model-00008-of-00015.safetensors",
    "model.encoder.layers.13.mlp.router.layer.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.13.mlp.router.top_k_bias": "model-00008-of-00015.safetensors",
    "model.encoder.layers.13.post_self_attn_layernorm.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.pre_self_attn_layernorm.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.self_attn.k_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.self_attn.o_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.self_attn.q_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.13.self_attn.v_proj.weight": "model-00007-of-00015.safetensors",
    "model.encoder.layers.14.mlp.experts.bias": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.mlp.experts.mlp.v1": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.mlp.experts.mlp.w1": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.mlp.experts.mlp.w2": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.mlp.router.layer.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.mlp.router.top_k_bias": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.post_self_attn_layernorm.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.pre_self_attn_layernorm.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.self_attn.k_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.self_attn.o_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.self_attn.q_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.14.self_attn.v_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.mlp.experts.bias": "model-00009-of-00015.safetensors",
    "model.encoder.layers.15.mlp.experts.mlp.v1": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.mlp.experts.mlp.w1": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.mlp.experts.mlp.w2": "model-00009-of-00015.safetensors",
    "model.encoder.layers.15.mlp.router.layer.weight": "model-00009-of-00015.safetensors",
    "model.encoder.layers.15.mlp.router.top_k_bias": "model-00009-of-00015.safetensors",
    "model.encoder.layers.15.post_self_attn_layernorm.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.pre_self_attn_layernorm.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.self_attn.k_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.self_attn.o_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.self_attn.q_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.15.self_attn.v_proj.weight": "model-00008-of-00015.safetensors",
    "model.encoder.layers.2.mlp.experts.bias": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.mlp.experts.mlp.v1": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.mlp.experts.mlp.w1": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.mlp.experts.mlp.w2": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.mlp.router.layer.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.mlp.router.top_k_bias": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.post_self_attn_layernorm.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.pre_self_attn_layernorm.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.self_attn.o_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.mlp.experts.bias": "model-00003-of-00015.safetensors",
    "model.encoder.layers.3.mlp.experts.mlp.v1": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.mlp.experts.mlp.w1": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.mlp.experts.mlp.w2": "model-00003-of-00015.safetensors",
    "model.encoder.layers.3.mlp.router.layer.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.3.mlp.router.top_k_bias": "model-00003-of-00015.safetensors",
    "model.encoder.layers.3.post_self_attn_layernorm.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.pre_self_attn_layernorm.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.self_attn.o_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00015.safetensors",
    "model.encoder.layers.4.mlp.experts.bias": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.mlp.experts.mlp.v1": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.mlp.experts.mlp.w1": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.mlp.experts.mlp.w2": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.mlp.router.layer.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.mlp.router.top_k_bias": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.post_self_attn_layernorm.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.pre_self_attn_layernorm.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.self_attn.k_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.self_attn.o_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.self_attn.q_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.4.self_attn.v_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.mlp.experts.bias": "model-00004-of-00015.safetensors",
    "model.encoder.layers.5.mlp.experts.mlp.v1": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.mlp.experts.mlp.w1": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.mlp.experts.mlp.w2": "model-00004-of-00015.safetensors",
    "model.encoder.layers.5.mlp.router.layer.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.5.mlp.router.top_k_bias": "model-00004-of-00015.safetensors",
    "model.encoder.layers.5.post_self_attn_layernorm.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.pre_self_attn_layernorm.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.self_attn.k_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.self_attn.o_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.self_attn.q_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.5.self_attn.v_proj.weight": "model-00003-of-00015.safetensors",
    "model.encoder.layers.6.mlp.experts.bias": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.mlp.experts.mlp.v1": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.mlp.experts.mlp.w1": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.mlp.experts.mlp.w2": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.mlp.router.layer.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.mlp.router.top_k_bias": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.post_self_attn_layernorm.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.pre_self_attn_layernorm.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.self_attn.k_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.self_attn.o_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.self_attn.q_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.6.self_attn.v_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.mlp.experts.bias": "model-00005-of-00015.safetensors",
    "model.encoder.layers.7.mlp.experts.mlp.v1": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.mlp.experts.mlp.w1": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.mlp.experts.mlp.w2": "model-00005-of-00015.safetensors",
    "model.encoder.layers.7.mlp.router.layer.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.7.mlp.router.top_k_bias": "model-00005-of-00015.safetensors",
    "model.encoder.layers.7.post_self_attn_layernorm.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.pre_self_attn_layernorm.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.self_attn.k_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.self_attn.o_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.self_attn.q_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.7.self_attn.v_proj.weight": "model-00004-of-00015.safetensors",
    "model.encoder.layers.8.mlp.experts.bias": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.mlp.experts.mlp.v1": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.mlp.experts.mlp.w1": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.mlp.experts.mlp.w2": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.mlp.router.layer.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.mlp.router.top_k_bias": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.post_self_attn_layernorm.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.pre_self_attn_layernorm.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.self_attn.k_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.self_attn.o_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.self_attn.q_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.8.self_attn.v_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.mlp.experts.bias": "model-00006-of-00015.safetensors",
    "model.encoder.layers.9.mlp.experts.mlp.v1": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.mlp.experts.mlp.w1": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.mlp.experts.mlp.w2": "model-00006-of-00015.safetensors",
    "model.encoder.layers.9.mlp.router.layer.weight": "model-00006-of-00015.safetensors",
    "model.encoder.layers.9.mlp.router.top_k_bias": "model-00006-of-00015.safetensors",
    "model.encoder.layers.9.post_self_attn_layernorm.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.pre_self_attn_layernorm.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.self_attn.k_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.self_attn.o_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.self_attn.q_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.layers.9.self_attn.v_proj.weight": "model-00005-of-00015.safetensors",
    "model.encoder.norm.weight": "model-00001-of-00015.safetensors"
  }
}