Initialize Llama3.1 with Whisper Tokenizer

config.json

@@ -35,5 +35,5 @@
   "torch_dtype": "bfloat16",
   "transformers_version": "4.43.1",
   "use_cache": true,
-  "vocab_size": 128770
+  "vocab_size": 128771
 }

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd93f43c4431d6b3bfbb623b375df0ed8fa8d398deb19d2e1048334ada622de0
-size 4980909360
+oid sha256:e185eae1705d58503daf684339d5d5838c61d86fbc6f421dd7a9471d084b38a4
+size 4980917552

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f135307bdfdedab660a818def328d63d81d1ff0b6b6992ccc9a6988efe885d9
-size 1172349496
+oid sha256:e84166fbae2e692893ba5773474a2f5bfa37e9839f32c19a8725d33e10abffa7
+size 1172357688

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
   "metadata": {
-    "total_size": 16068943872
+    "total_size": 16068960256
   },
   "weight_map": {
     "lm_head.weight": "model-00004-of-00004.safetensors",

tokenizer.json

@@ -6932,6 +6932,15 @@
       "rstrip": false,
       "normalized": false,
       "special": true
+    },
+    {
+      "id": 128770,
+      "content": "<|sound_0512|>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
     }
   ],
   "normalizer": null,

tokenizer_config.json

@@ -6159,6 +6159,14 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "128770": {
+      "content": "<|sound_0512|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
   "bos_token": "<|begin_of_text|>",