{ "metadata": { "total_size": 7006670800 }, "weight_map": { "lm_head.weight": "model-00002-of-00002.safetensors", "model.embed_tokens.weight": "model-00001-of-00002.safetensors", "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.0.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.0.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.0.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.0.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.0.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.0.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.0.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.1.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.1.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.1.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.1.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.1.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.1.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.1.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.10.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.10.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.10.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.10.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.10.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.10.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.10.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.11.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.11.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.11.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.11.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.11.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.11.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.11.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.12.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.12.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.12.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.12.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.12.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.12.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.12.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.13.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.13.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.13.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.13.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.13.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.13.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.13.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.14.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.14.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.14.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.14.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.14.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.14.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.14.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.15.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.15.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.15.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.15.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.15.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.15.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.15.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.16.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.16.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.16.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.16.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.16.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.16.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.16.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.17.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.17.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.17.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.17.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.17.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.17.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.17.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.18.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.18.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.18.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.18.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.18.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.18.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.18.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.19.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.19.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.19.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.19.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.19.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.19.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.19.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.2.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.2.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.2.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.2.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.2.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.2.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.2.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.20.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.20.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.20.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.20.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.20.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.20.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.20.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.21.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.21.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.21.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.21.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.21.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.21.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.21.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.22.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.22.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.22.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.22.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.22.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.22.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.22.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.23.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.23.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.23.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.23.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.23.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.23.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.23.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.24.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.24.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.24.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.24.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.24.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.24.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.24.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.25.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.25.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.25.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.25.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.25.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.25.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.25.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.26.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.26.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.26.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.26.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.26.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.26.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.26.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.27.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.27.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.27.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.27.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.27.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.27.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.27.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.28.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.28.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.28.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.28.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.28.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.28.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.28.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.29.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.29.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.29.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.29.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.29.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.29.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.29.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.29.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.3.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.3.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.3.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.3.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.3.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.3.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.3.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.30.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.30.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.30.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.30.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.30.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.30.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.30.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.31.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.31.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.31.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.31.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.31.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.31.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.31.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.32.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.32.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.32.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.32.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.32.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.32.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.32.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.33.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.33.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.33.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.33.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.33.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.33.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.33.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.34.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.34.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.34.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.34.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.34.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.34.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.34.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.35.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.35.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.35.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.35.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.35.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.35.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.35.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.36.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.36.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.36.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.36.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.36.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.36.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.36.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.36.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.36.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.37.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.37.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.37.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.37.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.37.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.37.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.37.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.37.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.37.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.38.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.38.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.38.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.38.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.38.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.38.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.38.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.38.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.38.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.39.input_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.39.mlp.down_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.39.mlp.down_proj.SU": "model-00002-of-00002.safetensors", "model.layers.39.mlp.down_proj.SV": "model-00002-of-00002.safetensors", "model.layers.39.mlp.down_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.39.mlp.down_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.SU": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.SV": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.39.mlp.upgate_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.39.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.o_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.o_proj.SU": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.o_proj.SV": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.o_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.o_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.Qidxs": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.SU": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.SV": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.Wscale": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.codebook_id": "model-00002-of-00002.safetensors", "model.layers.39.self_attn.qkv_proj.fuse_scales": "model-00002-of-00002.safetensors", "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.4.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.4.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.4.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.4.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.4.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.4.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.4.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.5.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.5.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.5.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.5.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.5.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.5.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.5.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.6.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.6.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.6.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.6.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.6.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.6.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.6.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.7.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.7.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.7.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.7.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.7.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.7.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.7.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.8.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.8.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.8.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.8.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.8.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.8.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.8.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.9.mlp.down_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.9.mlp.down_proj.SU": "model-00001-of-00002.safetensors", "model.layers.9.mlp.down_proj.SV": "model-00001-of-00002.safetensors", "model.layers.9.mlp.down_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.9.mlp.down_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.SU": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.SV": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.9.mlp.upgate_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.o_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.o_proj.SU": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.o_proj.SV": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.o_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.o_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.Qidxs": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.SU": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.SV": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.Wscale": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.codebook_id": "model-00001-of-00002.safetensors", "model.layers.9.self_attn.qkv_proj.fuse_scales": "model-00001-of-00002.safetensors", "model.norm.weight": "model-00002-of-00002.safetensors" } }