Fix quantized weights

config.json

@@ -28,8 +28,7 @@
     "bits": 4,
     "group_size": 128,
     "modules_to_not_convert": [
-      "lm_head",
-      "gate"
+      "lm_head"
     ],
     "quant_method": "awq",
     "version": "gemm",
@@ -44,7 +43,7 @@
   },
   "rope_theta": 1000000.0,
   "tie_word_embeddings": false,
-  "torch_dtype": "bfloat16",
+  "torch_dtype": "float16",
   "transformers_version": "4.43.1",
   "use_cache": false,
   "vocab_size": 102400

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:483683fa524604961ed6e6314b55f60da34c32eb3377ee1a7acbc4b2ace3788b
-size 4996470920
+oid sha256:d2e439ec16548f043be8f92ff8a959a41b7982f5131e37a4a56bdb311e0a802d
+size 4996470888

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c06c4a4c07c06747cc1ddde7b96a00ca0a74c5db4f9c387cbe16ee54a6573e69
-size 4953084616
+oid sha256:4a9b242bdeb878fa05571ccea91a068ff34b03595303a0a3d9d9712c6df4696d
+size 4953084576

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd94bc7e5c68d4d29dd85c8868747793e658a3b3040d74ae9b16fdefe5eb10b5
-size 4953084624
+oid sha256:072155109bdd3ee825718e328aa0aec6cfc1beed0e1399ad1182f6f929b8d2b2
+size 4953084584

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ad5fcbc7dc95aad1a78a69cb81dde8b15721292869a52c973a2e8319c6cbdce
-size 3277554616
+oid sha256:9072b470507f5ddf9ac9ad0cba3747bf3796e4f979cfc444804051624e67a146
+size 3277554600