Modify chat_template.jinja

chat_template.jinja

@@ -2,16 +2,8 @@
 {%- set ns_vid = namespace(count=0) %}
 {%- if tools %}
     {{- '<|im_start|>system\n' }}
-    {%- if messages[0].role == 'system' %}
-        {%- if messages[0].content is string %}
-            {{- messages[0].content + '\n\n' }}
-        {%- elif messages[0].content is sequence %}
-            {%- for content_part in messages[0].content %}
-                {%- if content_part.type == 'text' %}
-                    {{- content_part.text + '\n\n' }}
-                {%- endif %}
-            {%- endfor %}
-        {%- endif %}
+    {%- if messages[0].role == 'system' and messages[0].content is string %}
+        {{- messages[0].content + '\n\n' }}
     {%- endif %}
     {{- '# Tools\n\n' }}
     {{- 'You may call one or more functions to assist with the user query.\n\n' }}
@@ -31,17 +23,7 @@
     {{- '</tool_call><|im_end|>\n' }}
 {%- else %}
     {%- if messages[0].role == 'system' %}
-        {{- '<|im_start|>system\n' }}
-        {%- if messages[0].content is string %}
-            {{- messages[0].content }}
-        {%- elif messages[0].content is sequence %}
-            {%- for content_part in messages[0].content %}
-                {%- if content_part.type == 'text' %}
-                    {{- content_part.text }}
-                {%- endif %}
-            {%- endfor %}
-        {%- endif %}
-        {{- '<|im_end|>\n' }}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
     {%- endif %}
 {%- endif %}
 {%- set ns = namespace(last_user_index=-1) %}
@@ -51,19 +33,9 @@
     {%- endif %}
 {%- endfor %}
 {%- for message in messages %}
-    {%- set content = message.content %}
+    {%- set content = message.get('content', '') or '' %}
     {%- if (message.role == 'system' and not loop.first) %}
-        {{- '<|im_start|>' + message.role + '\n' }}
-        {%- if content is string %}
-            {{- content }}
-        {%- elif content is sequence %}
-            {%- for content_part in content %}
-                {%- if content_part.type == 'text' %}
-                    {{- content_part.text }}
-                {%- endif %}
-            {%- endfor %}
-        {%- endif %}
-        {{- '<|im_end|>' + '\n' }}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>\n' }}
     {%- elif message.role == 'user' %}
         {{- '<|im_start|>user\n' }}
         {%- if message['content'] is string %}
@@ -73,17 +45,30 @@
                 {%- if not loop.first %}
                     {{- '\n' }}
                 {%- endif %}
-                {%- if content['type'] == 'image' %}
-                    {%- set image_id = 'image_%02d' % ns_img.count %}
-                    {%- set ns_img.count = ns_img.count + 1 %}
-                    {{- '<|mime_start|>{"id": "' + image_id + '", "type": "image/jpeg", "filename": "' + content.get('filename', "a.jpg") + '"}<|mime_end|>\n' }}
-                    {{- '<|image_start|><|IMAGE_PAD|><|image_end|>' }}
-                {%- elif content['type'] == 'video' %}
-                    {%- set video_id = 'video_%02d' % ns_vid.count %}
-                    {%- set ns_vid.count = ns_vid.count + 1 %}
-                    {{- '<|mime_start|>{"id": "' + video_id + '", "type": "video/mp4", "filename": "' + content.get('filename', "a.mp4") + '"}<|mime_end|>\n' }}
-                    {{- '<|video_aux_start|>다음 중 video_duration은 비디오 길이 정보입니다. 참고하여 답변하세요. {"video_duration": ' + (content.get('video_duration') | tojson if content.get('video_duration') else '<|video_duration|>') + '}<|video_aux_end|>\n'}}
-                    {{- '<|video_start|><|VIDEO_PAD|><|video_end|>\n'}}
+                {%- if content['type'] == 'image_url' %}
+                    {%- set media_url = content.get('image_url', {}).get('url', '') %}
+                    {%- set url_lower = media_url.lower() %}
+                    {%- set video_extensions = ['.mp4', '.avi', '.mov', '.mkv', '.webm', '.flv', '.wmv', '.m4v'] %}
+                    {%- set ns_check = namespace(is_video=False) %}
+                    {%- for ext in video_extensions %}
+                        {%- if url_lower.endswith(ext) %}
+                            {%- set ns_check.is_video = True %}
+                        {%- endif %}
+                    {%- endfor %}
+                    {%- if ns_check.is_video %}
+                        {%- set video_id = 'video_%02d' % ns_vid.count %}
+                        {%- set ns_vid.count = ns_vid.count + 1 %}
+                        {%- set filename = media_url.split('/')[-1] %}
+                        {{- '<|mime_start|>{"id": "' + video_id + '", "type": "video/mp4", "filename": "' + filename + '"}<|mime_end|>\n' }}
+                        {{- '<|video_aux_start|>다음 중 video_duration은 비디오 길이 정보입니다. 참고하여 답변하세요. {"video_duration": "<|video_meta_duration|>"}<|video_aux_end|>\n'}}
+                        {{- '<|video_start|><|VIDEO_PAD|><|video_end|>\n'}}
+                    {%- else %}
+                        {%- set image_id = 'image_%02d' % ns_img.count %}
+                        {%- set ns_img.count = ns_img.count + 1 %}
+                        {%- set filename = media_url.split('/')[-1] %}
+                        {{- '<|mime_start|>{"id": "' + image_id + '", "type": "image/jpeg", "filename": "' + filename + '"}<|mime_end|>\n' }}
+                        {{- '<|image_start|><|IMAGE_PAD|><|image_end|>' }}
+                    {%- endif %}
                 {%- elif content['type'] == 'text' %}
                     {{- content['text'] }}
                 {%- endif %}
@@ -92,8 +77,8 @@
         {%- endif %}
     {%- elif message.role == 'assistant' %}
         {%- set reasoning_content = '' %}
-        {%- if message.reasoning_content is string %}
-            {%- set reasoning_content = message.reasoning_content %}
+        {%- if message.get('reasoning_content') is string %}
+            {%- set reasoning_content = message.get('reasoning_content') %}
         {%- else %}
             {%- if '</think>' in content %}
                 {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
@@ -109,20 +94,24 @@
         {%- else %}
             {{- '<|im_start|>' + message.role + '\n' + content }}
         {%- endif %}
-        {%- if message.tool_calls %}
-            {%- for tool_call in message.tool_calls %}
+        {%- if message.get('tool_calls') %}
+            {%- for tool_call in message.get('tool_calls', []) %}
                 {%- if not loop.first or content %}
                     {{- '\n' }}
                 {%- endif %}
-                {%- if tool_call.function %}
-                    {%- set tool_call = tool_call.function %}
+                {%- if tool_call.get('function') %}
+                    {%- set tool_call = tool_call.get('function') %}
+                {%- endif %}
+                {{- '<tool_call>' + tool_call.get('name', '') + '\n' }}
+                {%- set _args = tool_call.get('arguments', {}) %}
+                {%- if _args is string %}
+                    {{- '<arguments>' + _args + '</arguments>\n' }}
+                {%- elif _args is mapping %}
+                    {%- for k, v in _args.items() %}
+                        {{- '<arg_key>' + k + '</arg_key>\n' }}
+                        {{- '<arg_value>' + (v | tojson(ensure_ascii=False) if v is not string else v) + '</arg_value>\n' }}
+                    {%- endfor %}
                 {%- endif %}
-                {{- '<tool_call>' + tool_call.name + '\n' }}
-                {%- set _args = tool_call.arguments %}
-                {%- for k, v in _args.items() %}
-                    {{- '<arg_key>' + k + '</arg_key>\n' }}
-                    {{- '<arg_value>' + (v | tojson(ensure_ascii=False) if v is not string else v) + '</arg_value>\n' }}
-                {%- endfor %}
                 {{- '</tool_call>' }}
             {%- endfor %}
         {%- endif %}
@@ -140,8 +129,9 @@
     {%- endif %}
 {%- endfor %}
 {%- if add_generation_prompt %}
-    {{- '<|im_start|>assistant\n<think>\n' }}
-    {%- if skip_reasoning is defined and skip_reasoning is true %}
-        {{- '\n</think>\n\n' }}
+    {%- if thinking is defined and thinking is true %}
+        {{- '<|im_start|>assistant\n<think>\n' }}
+    {%- else %}
+        {{- '<|im_start|>assistant\n<think>\n\n</think>\n\n' }}
     {%- endif %}
-{%- endif %}
+{%- endif %}