Update readme, chat_template

README.md

@@ -28,7 +28,7 @@ tags:
     <img src="https://img.shields.io/badge/Seed-Project Page-yellow"></a>
   <a href="https://github.com/ByteDance-Seed/seed-oss">
     <img src="https://img.shields.io/badge/Seed-Tech Report Coming Soon-red"></a>
-  <a href="https://huggingface.co/ByteDance-Seed">
+  <a href="https://huggingface.co/collections/ByteDance-Seed/seed-oss-68a609f4201e788db05b5dcd">
     <img src="https://img.shields.io/badge/Seed-Hugging Face-orange"></a>
   <br>
   <a href="./LICENSE">
@@ -36,7 +36,7 @@ tags:
 </p>
 
 > [!NOTE]
-> This model card is dedicated to the `Seed-OSS-36B-Instruct` model.
+> This model card is dedicated to the `Seed-OSS-36B-Base-Instruct` model.
 
 ## News
 - [2025/08/20]🔥We release `Seed-OSS-36B-Base` (both with and without synthetic data versions) and `Seed-OSS-36B-Instruct`.
@@ -312,12 +312,12 @@ Incorporating synthetic instruction data into pretraining leads to improved perf
 </tr>
 <tr>
 <td align="center">ArcAGI V2</td>
-<td align="center">50.3</td>
-<td align="center"><b>41.7</b></td>
-<td align="center">37.8</td>
-<td align="center">14.4</td>
+<td align="center">1.16</td>
+<td align="center"><b>1.74</b></td>
+<td align="center">0.87</td>
+<td align="center">0</td>
 <td align="center">-</td>
-<td align="center"><ins>40.6</ins></td>
+<td align="center"><ins>1.45</ins></td>
 </tr>
 <tr>
 <td align="center">KORBench</td>
@@ -463,6 +463,8 @@ Incorporating synthetic instruction data into pretraining leads to improved perf
 </sup><br/><sup>
 - The results of Gemma3-27B are sourced directly from its technical report.
 </sup><br/><sup>
+- The results of ArcAGI-V2 were measured on the official evaluation set, which was not involved in the training process.
+</sup><br/><sup>
 - Generation configs for Seed-OSS-36B-Instruct: temperature=1.1, top_p=0.95. Specifically, for Taubench, temperature=1, top_p=0.7.
 </sup><br/><sup>
 </sup>
@@ -474,7 +476,7 @@ Incorporating synthetic instruction data into pretraining leads to improved perf
 
 Users can flexibly specify the model's thinking budget. The figure below shows the performance curves across different tasks as the thinking budget varies. For simpler tasks (such as IFEval), the model's chain of thought (CoT) is shorter, and the score exhibits fluctuations as the thinking budget increases. For more challenging tasks (such as AIME and LiveCodeBench), the model's CoT is longer, and the score improves with an increase in the thinking budget.
 
-![thinking_budget](./thinking_budget.png)
+![thinking_budget](./figures/thinking_budget.png)
 
 Here is an example with a thinking budget set to 512: during the reasoning process, the model periodically triggers self-reflection to estimate the consumed and remaining budget, and delivers the final response once the budget is exhausted or the reasoning concludes.
 ```
@@ -567,7 +569,7 @@ Use vllm >= 0.10.0 or higher for inference.
 
 - First install vLLM with Seed-OSS support version:
 ```shell
-VLLM_USE_PRECOMPILED=1 VLLM_TEST_USE_PRECOMPILED_NIGHTLY_WHEEL=1 pip install git+ssh://git@github.com/FoolPlayer/vllm.git@seed-oss
+VLLM_USE_PRECOMPILED=1 VLLM_TEST_USE_PRECOMPILED_NIGHTLY_WHEEL=1 pip install git+https://github.com/vllm-project/vllm.git
 ```
 
 - Start vLLM API server:

chat_template.jinja

@@ -21,25 +21,25 @@
      8192:   1024,
      16384:  1024
 } -%}
-{# 找到 “大于等于 thinking_budget” 的第一个档位 #}
+{# Find the first gear that is greater than or equal to the thinking_budget. #}
 {%- set ns = namespace(interval = None) -%}
 {%- for k, v in budget_reflections_v05 | dictsort -%}
     {%- if ns.interval is none and thinking_budget <= k -%}
         {%- set ns.interval = v -%}
     {%- endif -%}
 {%- endfor -%}
-{# 若超过最大档位，则用最后一个档位的值 #}
+{# If it exceeds the maximum gear, use the value of the last gear #}
 {%- if ns.interval is none -%}
     {%- set ns.interval = budget_reflections_v05[16384] -%}
 {%- endif -%}
-{# ---------- 预处理 system 消息 ---------- #}
+{# ---------- Preprocess the system message ---------- #}
 {%- if messages[0]["role"] == "system" %}
 {%- set system_message = messages[0]["content"] %}
 {%- set loop_messages = messages[1:] %}
 {%- else %}
 {%- set loop_messages = messages %}
 {%- endif %}
-{# ---------- 确保 tools 存在 ---------- #}
+{# ---------- Ensure tools exist ---------- #}
 {%- if not tools is defined or tools is none %}
 {%- set tools = [] %}
 {%- endif %}
@@ -51,7 +51,7 @@
     {%- elif t == "array" -%}list
     {%- else -%}Any{%- endif -%}
 {%- endmacro -%}
-{# ---------- 输出 system 块 ---------- #}
+{# ---------- Output the system block ---------- #}
 {%- if system_message is defined %}
 {{ bos_token + "system\n" + system_message }}
 {%- else %}
@@ -105,7 +105,7 @@ def {{ item.function.name }}(
 
 {{"工具调用请遵循如下格式:\n<seed:tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>value_1</parameter>\n<parameter=example_parameter_2>This is the value for the second parameter\nthat can span\nmultiple lines</parameter>\n</function>\n</seed:tool_call>\n"}}
 {%- endif %}
-{# 结束 system 块行尾 #}
+{# End the system block line #}
 {%- if system_message is defined or tools is iterable and tools | length > 0 %}
 {{ eos_token }}
 {%- endif %}
@@ -121,7 +121,7 @@ def {{ item.function.name }}(
 {{ eos_token }}
 {%- endif %}
 {%- endif %}
-{# ---------- 逐条写出历史消息 ---------- #}
+{# ---------- List the historical messages one by one ---------- #}
 {%- for message in loop_messages %}
 {%- if message.role == "assistant"
   and message.tool_calls is defined
@@ -157,15 +157,15 @@ def {{ item.function.name }}(
 {%- if message.content is defined and message.content is string and message.content | trim | length > 0 %}
 {{ "\n" + message.content | trim + eos_token }}
 {%- endif %}
-{# 包括 tool 角色，在这个逻辑 #}
+{# Include the tool role #}
 {%- else %}
 {{ bos_token + message.role + "\n" + message.content + eos_token }}
 {%- endif %}
 {%- endfor %}
-{# ---------- 控制模型开始续写 ---------- #}
+{# ---------- Control the model to start continuation ---------- #}
 {%- if add_generation_prompt %}
 {{ bos_token+"assistant\n" }}
 {%- if thinking_budget == 0 %}
-{{ think_begin_token+budget_begin_token }}
+{{ think_begin_token + "\n" + budget_begin_token + "The current thinking budget is 0, so I will directly start answering the question." + budget_end_token + "\n" + think_end_token }}
 {%- endif %}
 {%- endif %}