update

added_tokens.json

@@ -8,6 +8,7 @@
   "<img>": 92544,
   "<quad>": 92547,
   "<ref>": 92549,
+  "<video>": 92553,
   "<|action_end|>": 92540,
   "<|action_start|>": 92541,
   "<|im_end|>": 92542,

chat_template.jinja

@@ -1,5 +1,5 @@
 {% for message in messages %}{{'<|im_start|>' + message['role'] + '
-'}}{% if message['content'] is string %}{{ message['content'] }}{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' %}{{ '<image>
+'}}{% if message['content'] is string %}{{ message['content'] }}{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' %}{{ '<IMG_CONTEXT>
 ' }}{% elif content['type'] == 'video' %}{{ '<video>
 ' }}{% elif content['type'] == 'text' %}{{ content['text'] }}{% endif %}{% endfor %}{% endif %}{{'<|im_end|>
 '}}{% endfor %}{% if add_generation_prompt %}{{'<|im_start|>assistant

processor_config.json

@@ -1,6 +1,4 @@
 {
-  "fake_image_token": "<image>",
-  "fake_video_token": "<video>",
   "image_seq_length": 256,
   "processor_class": "InternVLProcessor"
 }

special_tokens_map.json

@@ -23,8 +23,6 @@
     "rstrip": false,
     "single_word": false
   },
-  "context_image_token": "<IMG_CONTEXT>",
-  "end_image_token": "</img>",
   "eos_token": {
     "content": "</s>",
     "lstrip": false,
@@ -39,12 +37,12 @@
     "rstrip": false,
     "single_word": false
   },
-  "start_image_token": "<img>",
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
     "single_word": false
-  }
+  },
+  "video_token": "<video>"
 }

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:14cd1534c31a36e6d06178cf0561501f8637931f5da4ce70c882827ec34b9ed8
-size 10578358
+oid sha256:ab3c9a9a9b96c4b6c042fd747c62d971b6b3475e6d6804ac77ce49c3e26876b7
+size 10578541

tokenizer_config.json

@@ -1634,6 +1634,14 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "92553": {
+      "content": "<video>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
   "additional_special_tokens": [
@@ -1665,9 +1673,7 @@
   "end_image_token": "</img>",
   "eos_token": "</s>",
   "extra_special_tokens": {
-    "context_image_token": "<IMG_CONTEXT>",
-    "end_image_token": "</img>",
-    "start_image_token": "<img>"
+    "video_token": "<video>"
   },
   "legacy": false,
   "model_max_length": 8192,
@@ -1679,5 +1685,6 @@
   "start_image_token": "<img>",
   "tokenizer_class": "LlamaTokenizer",
   "unk_token": "<unk>",
-  "use_default_system_prompt": false
+  "use_default_system_prompt": false,
+  "video_token": "<video>"
 }