Add files using upload-large-folder tool

config.json

@@ -34,9 +34,9 @@
     "output_router_logits": false,
     "partial_rotary_factor": 0.25,
     "quantization": {
-        "group_size": 64,
+        "group_size": 32,
         "bits": 4,
-        "mode": "affine",
+        "mode": "mxfp4",
         "model.embed_tokens": {
             "group_size": 64,
             "bits": 8,
@@ -72,6 +72,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.0.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.0.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.0.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.1.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -102,6 +117,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.1.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.1.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.1.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.2.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -132,6 +162,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.2.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.2.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.2.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.3.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -167,6 +212,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.3.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.3.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.3.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.4.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -197,6 +257,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.4.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.4.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.4.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.5.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -227,6 +302,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.5.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.5.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.5.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.6.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -257,6 +347,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.6.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.6.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.6.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.7.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -292,6 +397,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.7.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.7.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.7.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.8.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -322,6 +442,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.8.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.8.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.8.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.9.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -352,6 +487,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.9.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.9.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.9.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.10.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -382,6 +532,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.10.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.10.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.10.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.11.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -417,6 +582,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.11.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.11.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.11.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.12.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -447,6 +627,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.12.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.12.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.12.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.13.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -477,6 +672,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.13.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.13.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.13.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.14.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -507,6 +717,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.14.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.14.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.14.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.15.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -542,6 +767,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.15.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.15.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.15.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.16.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -572,6 +812,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.16.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.16.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.16.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.17.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -602,6 +857,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.17.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.17.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.17.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.18.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -632,6 +902,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.18.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.18.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.18.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.19.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -667,6 +952,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.19.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.19.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.19.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.20.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -697,6 +997,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.20.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.20.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.20.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.21.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -727,22 +1042,37 @@
             "bits": 4,
             "mode": "mxfp4"
         },
-        "model.layers.22.linear_attn.in_proj_qkvz": {
+        "model.layers.21.mlp.shared_expert.gate_proj": {
             "group_size": 64,
             "bits": 8,
             "mode": "affine"
         },
-        "model.layers.22.linear_attn.in_proj_ba": {
+        "model.layers.21.mlp.shared_expert.down_proj": {
             "group_size": 64,
             "bits": 8,
             "mode": "affine"
         },
-        "model.layers.22.linear_attn.out_proj": {
+        "model.layers.21.mlp.shared_expert.up_proj": {
             "group_size": 64,
             "bits": 8,
             "mode": "affine"
         },
-        "model.layers.22.mlp.switch_mlp.gate_proj": {
+        "model.layers.22.linear_attn.in_proj_qkvz": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.linear_attn.in_proj_ba": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.linear_attn.out_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.mlp.switch_mlp.gate_proj": {
             "group_size": 32,
             "bits": 4,
             "mode": "mxfp4"
@@ -757,6 +1087,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.22.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.23.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -792,6 +1137,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.23.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.23.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.23.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.24.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -822,6 +1182,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.24.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.24.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.24.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.25.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -852,6 +1227,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.25.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.25.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.25.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.26.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -882,6 +1272,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.26.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.26.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.26.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.27.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -917,6 +1322,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.27.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.27.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.27.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.28.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -947,6 +1367,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.28.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.28.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.28.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.29.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -977,6 +1412,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.29.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.29.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.29.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.30.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1007,6 +1457,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.30.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.30.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.30.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.31.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1042,6 +1507,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.31.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.31.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.31.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.32.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1072,6 +1552,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.32.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.32.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.32.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.33.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1102,6 +1597,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.33.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.33.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.33.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.34.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1132,6 +1642,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.34.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.34.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.34.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.35.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1167,6 +1692,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.35.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.35.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.35.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.36.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1197,6 +1737,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.36.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.36.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.36.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.37.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1227,6 +1782,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.37.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.37.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.37.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.38.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1257,6 +1827,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.38.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.38.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.38.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.39.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1292,6 +1877,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.39.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.39.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.39.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.40.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1322,6 +1922,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.40.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.40.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.40.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.41.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1352,6 +1967,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.41.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.41.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.41.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.42.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1382,6 +2012,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.42.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.42.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.42.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.43.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1417,6 +2062,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.43.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.43.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.43.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.44.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1447,6 +2107,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.44.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.44.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.44.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.45.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1477,6 +2152,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.45.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.45.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.45.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.46.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1507,6 +2197,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.46.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.46.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.46.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.47.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1542,6 +2247,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.47.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.47.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.47.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "lm_head": {
             "group_size": 64,
             "bits": 8,
@@ -1549,9 +2269,9 @@
         }
     },
     "quantization_config": {
-        "group_size": 64,
+        "group_size": 32,
         "bits": 4,
-        "mode": "affine",
+        "mode": "mxfp4",
         "model.embed_tokens": {
             "group_size": 64,
             "bits": 8,
@@ -1587,6 +2307,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.0.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.0.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.0.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.1.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1617,6 +2352,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.1.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.1.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.1.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.2.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1647,6 +2397,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.2.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.2.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.2.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.3.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1682,6 +2447,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.3.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.3.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.3.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.4.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1712,6 +2492,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.4.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.4.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.4.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.5.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1737,10 +2532,25 @@
             "bits": 4,
             "mode": "mxfp4"
         },
-        "model.layers.5.mlp.switch_mlp.down_proj": {
-            "group_size": 32,
-            "bits": 4,
-            "mode": "mxfp4"
+        "model.layers.5.mlp.switch_mlp.down_proj": {
+            "group_size": 32,
+            "bits": 4,
+            "mode": "mxfp4"
+        },
+        "model.layers.5.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.5.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.5.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
         },
         "model.layers.6.linear_attn.in_proj_qkvz": {
             "group_size": 64,
@@ -1772,6 +2582,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.6.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.6.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.6.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.7.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1807,6 +2632,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.7.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.7.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.7.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.8.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1837,6 +2677,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.8.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.8.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.8.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.9.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1867,6 +2722,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.9.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.9.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.9.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.10.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1897,6 +2767,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.10.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.10.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.10.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.11.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -1932,6 +2817,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.11.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.11.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.11.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.12.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1962,6 +2862,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.12.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.12.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.12.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.13.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -1992,6 +2907,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.13.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.13.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.13.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.14.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2022,6 +2952,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.14.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.14.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.14.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.15.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2057,6 +3002,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.15.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.15.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.15.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.16.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2087,6 +3047,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.16.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.16.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.16.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.17.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2117,6 +3092,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.17.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.17.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.17.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.18.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2147,6 +3137,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.18.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.18.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.18.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.19.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2182,6 +3187,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.19.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.19.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.19.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.20.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2212,6 +3232,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.20.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.20.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.20.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.21.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2242,6 +3277,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.21.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.21.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.21.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.22.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2272,6 +3322,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.22.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.22.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.23.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2307,6 +3372,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.23.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.23.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.23.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.24.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2337,6 +3417,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.24.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.24.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.24.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.25.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2367,6 +3462,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.25.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.25.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.25.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.26.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2397,6 +3507,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.26.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.26.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.26.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.27.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2432,6 +3557,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.27.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.27.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.27.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.28.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2462,6 +3602,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.28.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.28.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.28.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.29.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2492,6 +3647,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.29.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.29.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.29.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.30.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2522,6 +3692,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.30.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.30.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.30.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.31.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2557,6 +3742,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.31.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.31.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.31.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.32.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2587,6 +3787,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.32.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.32.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.32.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.33.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2617,6 +3832,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.33.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.33.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.33.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.34.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2647,6 +3877,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.34.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.34.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.34.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.35.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2682,6 +3927,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.35.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.35.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.35.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.36.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2712,6 +3972,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.36.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.36.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.36.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.37.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2742,6 +4017,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.37.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.37.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.37.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.38.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2772,6 +4062,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.38.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.38.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.38.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.39.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2807,6 +4112,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.39.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.39.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.39.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.40.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2837,6 +4157,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.40.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.40.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.40.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.41.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2867,6 +4202,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.41.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.41.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.41.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.42.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2897,6 +4247,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.42.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.42.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.42.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.43.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -2932,6 +4297,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.43.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.43.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.43.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.44.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2962,6 +4342,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.44.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.44.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.44.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.45.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -2992,6 +4387,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.45.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.45.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.45.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.46.linear_attn.in_proj_qkvz": {
             "group_size": 64,
             "bits": 8,
@@ -3022,6 +4432,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.46.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.46.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.46.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "model.layers.47.self_attn.q_proj": {
             "group_size": 64,
             "bits": 8,
@@ -3057,6 +4482,21 @@
             "bits": 4,
             "mode": "mxfp4"
         },
+        "model.layers.47.mlp.shared_expert.gate_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.47.mlp.shared_expert.down_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
+        "model.layers.47.mlp.shared_expert.up_proj": {
+            "group_size": 64,
+            "bits": 8,
+            "mode": "affine"
+        },
         "lm_head": {
             "group_size": 64,
             "bits": 8,

model-00001-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a19555cb848e8db762bcd0423f35fd9b6fa3e2ef0f9daea337f8a378051b949
-size 5136685826
+oid sha256:2024fb443cbc42991f1b0e599cc023050b194d5f09b3d87586464b48e8c34320
+size 5131653128

model-00002-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05d512b2820924fe14ae37f6c469d91147e8acbc8cf087ac5d3c7fcc476d9db2
-size 5354539238
+oid sha256:2b1b9256e4fd01e45ee40801340b8b7150047daf24822b943c81ee595e10d842
+size 5367924965

model-00003-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c63f680db9da22a4651ec48e31ee47385d08fafdb87460ee0a152e5210314b9f
-size 5109835354
+oid sha256:df106317f9a4285b347e2ed95d5a2f1615cd19a137764f70841c72b9b24eb5ce
+size 5358036471

model-00004-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:08458af2bee021aec677bd00f0ffcabd6c24cd111f7233d16b605c1f6176a5bb
-size 5367122254
+oid sha256:1f3d496f8e61aaf132e9bb7c03fb710e1089e452f035c833de9961d0dd0e137d
+size 5367925082

model-00005-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:397a531b233251105ce94cfef0f24712d16b3615791cb1da0e2d732760b4bb27
-size 5368422424
+oid sha256:41c7fc16070ef760ae03df095948a8e069e3f4bfc1d919658595f6b85ee3304c
+size 5106378225

model-00006-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f60cf313663a1ab19ba9f384bee4dceee9c82a90ea9693cab5ccdb3e615087d
-size 5352187259
+oid sha256:d673eac7cfed394f114f9831c337de195e669a52b18c44470e2a2e2ea8d6d8f5
+size 5367924994

model-00007-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f75b6cf275b49b7e26b762634e190109c6829320f915fbbfd30083564d3e3e6c
-size 5352767299
+oid sha256:3c28d7a730fbd540f0afa1f5a782d54b37c9ff34765601ade354a421635455be
+size 5363794885

model-00008-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fbdee11a1b04dd2669539ef1b7d73536923733a7a3a6b8900270a056c8477f5a
-size 5109835416
+oid sha256:d47a37e53e70835b633d1e044262dcd100aae75d4880d14ca9e7cea2385e9eac
+size 5367923842

model-00009-of-00009.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65acab9e2eea14867519ee414621656e5e17bc8a5fb63a34059fad88fa7159da
-size 1508185434
+oid sha256:255c562d60a4d76022fabf121f9a4e4f2e5d8658cc443610b0173261e86d1d2c
+size 1200623808

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
     "metadata": {
-        "total_size": 43659382272,
+        "total_size": 43631998464,
         "total_parameters": 79674388992
     },
     "weight_map": {
@@ -34,8 +34,6 @@
         "model.layers.0.mlp.shared_expert.up_proj.biases": "model-00001-of-00009.safetensors",
         "model.layers.0.mlp.shared_expert.up_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.0.mlp.shared_expert.up_proj.weight": "model-00001-of-00009.safetensors",
-        "model.layers.0.mlp.shared_expert_gate.biases": "model-00001-of-00009.safetensors",
-        "model.layers.0.mlp.shared_expert_gate.scales": "model-00001-of-00009.safetensors",
         "model.layers.0.mlp.shared_expert_gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.0.mlp.switch_mlp.down_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.0.mlp.switch_mlp.down_proj.weight": "model-00001-of-00009.safetensors",
@@ -68,8 +66,6 @@
         "model.layers.1.mlp.shared_expert.up_proj.biases": "model-00001-of-00009.safetensors",
         "model.layers.1.mlp.shared_expert.up_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.1.mlp.shared_expert.up_proj.weight": "model-00001-of-00009.safetensors",
-        "model.layers.1.mlp.shared_expert_gate.biases": "model-00001-of-00009.safetensors",
-        "model.layers.1.mlp.shared_expert_gate.scales": "model-00001-of-00009.safetensors",
         "model.layers.1.mlp.shared_expert_gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.1.mlp.switch_mlp.down_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.1.mlp.switch_mlp.down_proj.weight": "model-00001-of-00009.safetensors",
@@ -102,8 +98,6 @@
         "model.layers.10.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.10.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.10.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.10.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.10.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.10.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.10.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.10.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",
@@ -123,12 +117,10 @@
         "model.layers.11.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.11.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.11.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.11.mlp.switch_mlp.gate_proj.scales": "model-00003-of-00009.safetensors",
+        "model.layers.11.mlp.switch_mlp.gate_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.11.mlp.switch_mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
         "model.layers.11.mlp.switch_mlp.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.11.mlp.switch_mlp.up_proj.weight": "model-00003-of-00009.safetensors",
@@ -171,8 +163,6 @@
         "model.layers.12.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.12.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.12.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.12.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.12.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.12.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.12.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.12.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
@@ -205,8 +195,6 @@
         "model.layers.13.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.13.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.13.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.13.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.13.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.13.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.13.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.13.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
@@ -239,8 +227,6 @@
         "model.layers.14.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.14.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.14.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.14.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.14.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.14.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.14.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.14.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
@@ -260,8 +246,6 @@
         "model.layers.15.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.15.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.15.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.15.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.15.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.15.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.15.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.15.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
@@ -308,8 +292,6 @@
         "model.layers.16.mlp.shared_expert.up_proj.biases": "model-00003-of-00009.safetensors",
         "model.layers.16.mlp.shared_expert.up_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.16.mlp.shared_expert.up_proj.weight": "model-00003-of-00009.safetensors",
-        "model.layers.16.mlp.shared_expert_gate.biases": "model-00003-of-00009.safetensors",
-        "model.layers.16.mlp.shared_expert_gate.scales": "model-00003-of-00009.safetensors",
         "model.layers.16.mlp.shared_expert_gate.weight": "model-00003-of-00009.safetensors",
         "model.layers.16.mlp.switch_mlp.down_proj.scales": "model-00003-of-00009.safetensors",
         "model.layers.16.mlp.switch_mlp.down_proj.weight": "model-00003-of-00009.safetensors",
@@ -342,13 +324,11 @@
         "model.layers.17.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.17.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.17.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.switch_mlp.gate_proj.scales": "model-00004-of-00009.safetensors",
-        "model.layers.17.mlp.switch_mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
+        "model.layers.17.mlp.switch_mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
         "model.layers.17.mlp.switch_mlp.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.17.mlp.switch_mlp.up_proj.weight": "model-00004-of-00009.safetensors",
         "model.layers.17.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
@@ -376,8 +356,6 @@
         "model.layers.18.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.18.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.18.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.18.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.18.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.18.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.18.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.18.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
@@ -397,8 +375,6 @@
         "model.layers.19.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.19.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.19.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.19.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.19.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.19.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.19.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.19.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
@@ -445,8 +421,6 @@
         "model.layers.2.mlp.shared_expert.up_proj.biases": "model-00001-of-00009.safetensors",
         "model.layers.2.mlp.shared_expert.up_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.2.mlp.shared_expert.up_proj.weight": "model-00001-of-00009.safetensors",
-        "model.layers.2.mlp.shared_expert_gate.biases": "model-00001-of-00009.safetensors",
-        "model.layers.2.mlp.shared_expert_gate.scales": "model-00001-of-00009.safetensors",
         "model.layers.2.mlp.shared_expert_gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.2.mlp.switch_mlp.down_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.2.mlp.switch_mlp.down_proj.weight": "model-00001-of-00009.safetensors",
@@ -479,8 +453,6 @@
         "model.layers.20.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.20.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.20.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.20.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.20.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.20.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.20.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.20.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
@@ -513,8 +485,6 @@
         "model.layers.21.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.21.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.21.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.21.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.21.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.21.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.21.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.21.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
@@ -547,8 +517,6 @@
         "model.layers.22.mlp.shared_expert.up_proj.biases": "model-00004-of-00009.safetensors",
         "model.layers.22.mlp.shared_expert.up_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.22.mlp.shared_expert.up_proj.weight": "model-00004-of-00009.safetensors",
-        "model.layers.22.mlp.shared_expert_gate.biases": "model-00004-of-00009.safetensors",
-        "model.layers.22.mlp.shared_expert_gate.scales": "model-00004-of-00009.safetensors",
         "model.layers.22.mlp.shared_expert_gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.22.mlp.switch_mlp.down_proj.scales": "model-00004-of-00009.safetensors",
         "model.layers.22.mlp.switch_mlp.down_proj.weight": "model-00004-of-00009.safetensors",
@@ -558,7 +526,7 @@
         "model.layers.22.mlp.switch_mlp.up_proj.weight": "model-00004-of-00009.safetensors",
         "model.layers.22.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
         "model.layers.23.input_layernorm.weight": "model-00004-of-00009.safetensors",
-        "model.layers.23.mlp.gate.weight": "model-00005-of-00009.safetensors",
+        "model.layers.23.mlp.gate.weight": "model-00004-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert.down_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -568,13 +536,11 @@
         "model.layers.23.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.23.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.23.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.switch_mlp.gate_proj.scales": "model-00005-of-00009.safetensors",
-        "model.layers.23.mlp.switch_mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
+        "model.layers.23.mlp.switch_mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
         "model.layers.23.mlp.switch_mlp.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.23.mlp.switch_mlp.up_proj.weight": "model-00005-of-00009.safetensors",
         "model.layers.23.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
@@ -616,8 +582,6 @@
         "model.layers.24.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.24.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.24.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.24.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.24.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.24.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.24.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.24.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -650,8 +614,6 @@
         "model.layers.25.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.25.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.25.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.25.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.25.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.25.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.25.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.25.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -684,8 +646,6 @@
         "model.layers.26.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.26.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.26.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.26.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.26.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.26.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.26.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.26.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -705,8 +665,6 @@
         "model.layers.27.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.27.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.27.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.27.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.27.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.27.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.27.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.27.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -753,8 +711,6 @@
         "model.layers.28.mlp.shared_expert.up_proj.biases": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.shared_expert.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.shared_expert.up_proj.weight": "model-00005-of-00009.safetensors",
-        "model.layers.28.mlp.shared_expert_gate.biases": "model-00005-of-00009.safetensors",
-        "model.layers.28.mlp.shared_expert_gate.scales": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.shared_expert_gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.switch_mlp.down_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.switch_mlp.down_proj.weight": "model-00005-of-00009.safetensors",
@@ -763,21 +719,21 @@
         "model.layers.28.mlp.switch_mlp.up_proj.scales": "model-00005-of-00009.safetensors",
         "model.layers.28.mlp.switch_mlp.up_proj.weight": "model-00005-of-00009.safetensors",
         "model.layers.28.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
-        "model.layers.29.input_layernorm.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.A_log": "model-00006-of-00009.safetensors",
+        "model.layers.29.input_layernorm.weight": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.A_log": "model-00005-of-00009.safetensors",
         "model.layers.29.linear_attn.conv1d.weight": "model-00005-of-00009.safetensors",
-        "model.layers.29.linear_attn.dt_bias": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.in_proj_ba.biases": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.in_proj_ba.scales": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.in_proj_ba.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.in_proj_qkvz.biases": "model-00006-of-00009.safetensors",
+        "model.layers.29.linear_attn.dt_bias": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.in_proj_ba.biases": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.in_proj_ba.scales": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.in_proj_ba.weight": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.in_proj_qkvz.biases": "model-00005-of-00009.safetensors",
         "model.layers.29.linear_attn.in_proj_qkvz.scales": "model-00005-of-00009.safetensors",
         "model.layers.29.linear_attn.in_proj_qkvz.weight": "model-00005-of-00009.safetensors",
-        "model.layers.29.linear_attn.norm.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.out_proj.biases": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.out_proj.scales": "model-00006-of-00009.safetensors",
-        "model.layers.29.linear_attn.out_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.mlp.gate.weight": "model-00006-of-00009.safetensors",
+        "model.layers.29.linear_attn.norm.weight": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.out_proj.biases": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.out_proj.scales": "model-00005-of-00009.safetensors",
+        "model.layers.29.linear_attn.out_proj.weight": "model-00005-of-00009.safetensors",
+        "model.layers.29.mlp.gate.weight": "model-00005-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert.down_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -787,8 +743,6 @@
         "model.layers.29.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.29.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -796,7 +750,7 @@
         "model.layers.29.mlp.switch_mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.switch_mlp.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.29.mlp.switch_mlp.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.29.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
+        "model.layers.29.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
         "model.layers.3.input_layernorm.weight": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.shared_expert.down_proj.biases": "model-00001-of-00009.safetensors",
@@ -808,8 +762,6 @@
         "model.layers.3.mlp.shared_expert.up_proj.biases": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.shared_expert.up_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.shared_expert.up_proj.weight": "model-00001-of-00009.safetensors",
-        "model.layers.3.mlp.shared_expert_gate.biases": "model-00001-of-00009.safetensors",
-        "model.layers.3.mlp.shared_expert_gate.scales": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.shared_expert_gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.switch_mlp.down_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.3.mlp.switch_mlp.down_proj.weight": "model-00001-of-00009.safetensors",
@@ -856,8 +808,6 @@
         "model.layers.30.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.30.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.30.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.30.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.30.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.30.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.30.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.30.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -877,8 +827,6 @@
         "model.layers.31.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.31.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.31.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.31.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.31.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.31.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.31.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.31.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -925,8 +873,6 @@
         "model.layers.32.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.32.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.32.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.32.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.32.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.32.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.32.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.32.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -959,8 +905,6 @@
         "model.layers.33.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.33.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.33.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.33.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.33.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.33.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.33.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.33.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -993,8 +937,6 @@
         "model.layers.34.mlp.shared_expert.up_proj.biases": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.shared_expert.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.shared_expert.up_proj.weight": "model-00006-of-00009.safetensors",
-        "model.layers.34.mlp.shared_expert_gate.biases": "model-00006-of-00009.safetensors",
-        "model.layers.34.mlp.shared_expert_gate.scales": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.shared_expert_gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.switch_mlp.down_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.switch_mlp.down_proj.weight": "model-00006-of-00009.safetensors",
@@ -1003,8 +945,8 @@
         "model.layers.34.mlp.switch_mlp.up_proj.scales": "model-00006-of-00009.safetensors",
         "model.layers.34.mlp.switch_mlp.up_proj.weight": "model-00006-of-00009.safetensors",
         "model.layers.34.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
-        "model.layers.35.input_layernorm.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.mlp.gate.weight": "model-00007-of-00009.safetensors",
+        "model.layers.35.input_layernorm.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.mlp.gate.weight": "model-00006-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert.down_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1014,8 +956,6 @@
         "model.layers.35.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.mlp.shared_expert_gate.biases": "model-00007-of-00009.safetensors",
-        "model.layers.35.mlp.shared_expert_gate.scales": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1023,21 +963,21 @@
         "model.layers.35.mlp.switch_mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.switch_mlp.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.35.mlp.switch_mlp.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.k_norm.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.k_proj.biases": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.k_proj.scales": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.o_proj.biases": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.o_proj.scales": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.q_norm.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.q_proj.biases": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.q_proj.scales": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.v_proj.biases": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.v_proj.scales": "model-00007-of-00009.safetensors",
-        "model.layers.35.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
+        "model.layers.35.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.k_norm.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.k_proj.biases": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.k_proj.scales": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.o_proj.biases": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.o_proj.scales": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.q_norm.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.q_proj.biases": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.q_proj.scales": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.v_proj.biases": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.v_proj.scales": "model-00006-of-00009.safetensors",
+        "model.layers.35.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
         "model.layers.36.input_layernorm.weight": "model-00007-of-00009.safetensors",
         "model.layers.36.linear_attn.A_log": "model-00007-of-00009.safetensors",
         "model.layers.36.linear_attn.conv1d.weight": "model-00007-of-00009.safetensors",
@@ -1062,8 +1002,6 @@
         "model.layers.36.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.36.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.36.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.36.mlp.shared_expert_gate.biases": "model-00007-of-00009.safetensors",
-        "model.layers.36.mlp.shared_expert_gate.scales": "model-00007-of-00009.safetensors",
         "model.layers.36.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
         "model.layers.36.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.36.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1096,8 +1034,6 @@
         "model.layers.37.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.37.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.37.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.37.mlp.shared_expert_gate.biases": "model-00007-of-00009.safetensors",
-        "model.layers.37.mlp.shared_expert_gate.scales": "model-00007-of-00009.safetensors",
         "model.layers.37.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
         "model.layers.37.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.37.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1130,8 +1066,6 @@
         "model.layers.38.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.38.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.38.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.38.mlp.shared_expert_gate.biases": "model-00007-of-00009.safetensors",
-        "model.layers.38.mlp.shared_expert_gate.scales": "model-00007-of-00009.safetensors",
         "model.layers.38.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
         "model.layers.38.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.38.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1151,8 +1085,6 @@
         "model.layers.39.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
         "model.layers.39.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.39.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
-        "model.layers.39.mlp.shared_expert_gate.biases": "model-00007-of-00009.safetensors",
-        "model.layers.39.mlp.shared_expert_gate.scales": "model-00007-of-00009.safetensors",
         "model.layers.39.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
         "model.layers.39.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.39.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
@@ -1199,8 +1131,6 @@
         "model.layers.4.mlp.shared_expert.up_proj.biases": "model-00001-of-00009.safetensors",
         "model.layers.4.mlp.shared_expert.up_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.4.mlp.shared_expert.up_proj.weight": "model-00001-of-00009.safetensors",
-        "model.layers.4.mlp.shared_expert_gate.biases": "model-00001-of-00009.safetensors",
-        "model.layers.4.mlp.shared_expert_gate.scales": "model-00001-of-00009.safetensors",
         "model.layers.4.mlp.shared_expert_gate.weight": "model-00001-of-00009.safetensors",
         "model.layers.4.mlp.switch_mlp.down_proj.scales": "model-00001-of-00009.safetensors",
         "model.layers.4.mlp.switch_mlp.down_proj.weight": "model-00001-of-00009.safetensors",
@@ -1224,19 +1154,17 @@
         "model.layers.40.linear_attn.out_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.40.linear_attn.out_proj.weight": "model-00007-of-00009.safetensors",
         "model.layers.40.mlp.gate.weight": "model-00007-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.down_proj.biases": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.down_proj.scales": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.down_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.gate_proj.biases": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.gate_proj.scales": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.gate_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
-        "model.layers.40.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.down_proj.biases": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.down_proj.scales": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.down_proj.weight": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.gate_proj.biases": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.gate_proj.scales": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.gate_proj.weight": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.up_proj.biases": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.up_proj.scales": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert.up_proj.weight": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.shared_expert_gate.weight": "model-00007-of-00009.safetensors",
+        "model.layers.40.mlp.switch_mlp.down_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.40.mlp.switch_mlp.down_proj.weight": "model-00007-of-00009.safetensors",
         "model.layers.40.mlp.switch_mlp.gate_proj.scales": "model-00007-of-00009.safetensors",
         "model.layers.40.mlp.switch_mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
@@ -1244,16 +1172,16 @@
         "model.layers.40.mlp.switch_mlp.up_proj.weight": "model-00007-of-00009.safetensors",
         "model.layers.40.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
         "model.layers.41.input_layernorm.weight": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.A_log": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.conv1d.weight": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.dt_bias": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_ba.biases": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_ba.scales": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_ba.weight": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_qkvz.biases": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_qkvz.scales": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.in_proj_qkvz.weight": "model-00008-of-00009.safetensors",
-        "model.layers.41.linear_attn.norm.weight": "model-00008-of-00009.safetensors",
+        "model.layers.41.linear_attn.A_log": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.conv1d.weight": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.dt_bias": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_ba.biases": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_ba.scales": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_ba.weight": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_qkvz.biases": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_qkvz.scales": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.in_proj_qkvz.weight": "model-00007-of-00009.safetensors",
+        "model.layers.41.linear_attn.norm.weight": "model-00007-of-00009.safetensors",
         "model.layers.41.linear_attn.out_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.41.linear_attn.out_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.41.linear_attn.out_proj.weight": "model-00008-of-00009.safetensors",
@@ -1267,8 +1195,6 @@
         "model.layers.41.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.41.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.41.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.41.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.41.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
         "model.layers.41.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
         "model.layers.41.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.41.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
@@ -1301,8 +1227,6 @@
         "model.layers.42.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.42.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.42.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.42.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.42.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
         "model.layers.42.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
         "model.layers.42.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.42.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
@@ -1322,8 +1246,6 @@
         "model.layers.43.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.43.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.43.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.43.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.43.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
         "model.layers.43.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
         "model.layers.43.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.43.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
@@ -1370,8 +1292,6 @@
         "model.layers.44.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.44.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.44.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.44.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.44.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
         "model.layers.44.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
         "model.layers.44.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.44.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
@@ -1404,8 +1324,6 @@
         "model.layers.45.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
         "model.layers.45.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.45.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
-        "model.layers.45.mlp.shared_expert_gate.biases": "model-00008-of-00009.safetensors",
-        "model.layers.45.mlp.shared_expert_gate.scales": "model-00008-of-00009.safetensors",
         "model.layers.45.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
         "model.layers.45.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.45.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
@@ -1429,20 +1347,18 @@
         "model.layers.46.linear_attn.out_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.46.linear_attn.out_proj.weight": "model-00008-of-00009.safetensors",
         "model.layers.46.mlp.gate.weight": "model-00008-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.down_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.down_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.down_proj.weight": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.gate_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.gate_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.gate_proj.weight": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.up_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.up_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert.up_proj.weight": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert_gate.biases": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert_gate.scales": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.shared_expert_gate.weight": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.switch_mlp.down_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.46.mlp.switch_mlp.down_proj.weight": "model-00009-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.down_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.down_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.down_proj.weight": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.gate_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.gate_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.gate_proj.weight": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.up_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.up_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert.up_proj.weight": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.shared_expert_gate.weight": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.switch_mlp.down_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.46.mlp.switch_mlp.down_proj.weight": "model-00008-of-00009.safetensors",
         "model.layers.46.mlp.switch_mlp.gate_proj.scales": "model-00008-of-00009.safetensors",
         "model.layers.46.mlp.switch_mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
         "model.layers.46.mlp.switch_mlp.up_proj.scales": "model-00008-of-00009.safetensors",
@@ -1459,8 +1375,6 @@
         "model.layers.47.mlp.shared_expert.up_proj.biases": "model-00009-of-00009.safetensors",
         "model.layers.47.mlp.shared_expert.up_proj.scales": "model-00009-of-00009.safetensors",
         "model.layers.47.mlp.shared_expert.up_proj.weight": "model-00009-of-00009.safetensors",
-        "model.layers.47.mlp.shared_expert_gate.biases": "model-00009-of-00009.safetensors",
-        "model.layers.47.mlp.shared_expert_gate.scales": "model-00009-of-00009.safetensors",
         "model.layers.47.mlp.shared_expert_gate.weight": "model-00009-of-00009.safetensors",
         "model.layers.47.mlp.switch_mlp.down_proj.scales": "model-00009-of-00009.safetensors",
         "model.layers.47.mlp.switch_mlp.down_proj.weight": "model-00009-of-00009.safetensors",
@@ -1470,19 +1384,19 @@
         "model.layers.47.mlp.switch_mlp.up_proj.weight": "model-00009-of-00009.safetensors",
         "model.layers.47.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
         "model.layers.47.self_attn.k_norm.weight": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.k_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.k_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
+        "model.layers.47.self_attn.k_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.k_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
         "model.layers.47.self_attn.o_proj.biases": "model-00009-of-00009.safetensors",
         "model.layers.47.self_attn.o_proj.scales": "model-00009-of-00009.safetensors",
         "model.layers.47.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
         "model.layers.47.self_attn.q_norm.weight": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.q_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.q_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.v_proj.biases": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.v_proj.scales": "model-00009-of-00009.safetensors",
-        "model.layers.47.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
+        "model.layers.47.self_attn.q_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.q_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.v_proj.biases": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.v_proj.scales": "model-00008-of-00009.safetensors",
+        "model.layers.47.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
         "model.layers.5.input_layernorm.weight": "model-00001-of-00009.safetensors",
         "model.layers.5.linear_attn.A_log": "model-00001-of-00009.safetensors",
         "model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00009.safetensors",
@@ -1507,8 +1421,6 @@
         "model.layers.5.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.5.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.5.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.5.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.5.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.5.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.5.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.5.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",
@@ -1541,8 +1453,6 @@
         "model.layers.6.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.6.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.6.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.6.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.6.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.6.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.6.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.6.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",
@@ -1562,8 +1472,6 @@
         "model.layers.7.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.7.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.7.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.7.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.7.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.7.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.7.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.7.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",
@@ -1610,8 +1518,6 @@
         "model.layers.8.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.8.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.8.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.8.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.8.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.8.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.8.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.8.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",
@@ -1644,8 +1550,6 @@
         "model.layers.9.mlp.shared_expert.up_proj.biases": "model-00002-of-00009.safetensors",
         "model.layers.9.mlp.shared_expert.up_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.9.mlp.shared_expert.up_proj.weight": "model-00002-of-00009.safetensors",
-        "model.layers.9.mlp.shared_expert_gate.biases": "model-00002-of-00009.safetensors",
-        "model.layers.9.mlp.shared_expert_gate.scales": "model-00002-of-00009.safetensors",
         "model.layers.9.mlp.shared_expert_gate.weight": "model-00002-of-00009.safetensors",
         "model.layers.9.mlp.switch_mlp.down_proj.scales": "model-00002-of-00009.safetensors",
         "model.layers.9.mlp.switch_mlp.down_proj.weight": "model-00002-of-00009.safetensors",