Add files using upload-large-folder tool

.gitattributes

@@ -39,3 +39,5 @@ assets/architecture.png filter=lfs diff=lfs merge=lfs -text
 assets/logo.png filter=lfs diff=lfs merge=lfs -text
 assets/mossvl_quantization_benchmark_comparison_zh_4k.png filter=lfs diff=lfs merge=lfs -text
 assets/mossvl_quantization_benchmark_comparison_en_4k.png filter=lfs diff=lfs merge=lfs -text
+assets/mossvl_quantization_benchmark_comparison_final_v3_en_4k.png filter=lfs diff=lfs merge=lfs -text
+assets/mossvl_quantization_benchmark_comparison_final_v3_zh_4k.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -54,14 +54,15 @@ inference. This checkpoint is not an SGLang release.
 directly and do not pass a second quantization configuration or replace its
 generation config with the BF16 source file.
 
-## Base model benchmark
+## Quantization benchmark
 
-The following figure reports the streaming benchmark results of the original
-MOSS-VL-Realtime release. It provides base-model context; it is not presented
-as a separate quality evaluation of this FP8 checkpoint.
+The final evaluation compares the original BF16 model with all four release
+profiles on their corresponding benchmark suites. For this streaming FP8
+checkpoint, the scores are 70.66 on OVOBench Avg, 62.93 on StreamingBench Avg,
+and 65.50 on OmniMMI PA, compared with 70.86, 62.42, and 66.00 for BF16.
 
 <p align="center">
-  <img src="assets/benchmark-streaming.png" alt="MOSS-VL-Realtime benchmark" width="100%"/>
+  <img src="assets/mossvl_quantization_benchmark_comparison_final_v3_en_4k.png" alt="MOSS-VL quantization benchmark comparison" width="100%"/>
 </p>
 
 ## Hardware requirements

README_zh.md

@@ -52,13 +52,14 @@ tags:
 `generation_config.json` 会自动启用 HQQ KV8。请直接加载 checkpoint，
 不要再次传入量化配置，也不要用 BF16 原模型的 generation config 覆盖它。
 
-## 基础模型 Benchmark
+## 量化 Benchmark
 
-下图是原始 MOSS-VL-Realtime 的流式 benchmark，用于说明基础模型能力；
-它不作为该 FP8 checkpoint 的独立精度测评结果。
+最终测评在各自对应的 benchmark 上对比原始 BF16 模型与四个量化发布配置。
+该流式 FP8 checkpoint 的 OVOBench Avg、StreamingBench Avg 和 OmniMMI PA
+分别为 70.66、62.93 和 65.50；对应 BF16 分数为 70.86、62.42 和 66.00。
 
 <p align="center">
-  <img src="assets/benchmark-streaming.png" alt="MOSS-VL-Realtime benchmark" width="100%"/>
+  <img src="assets/mossvl_quantization_benchmark_comparison_final_v3_zh_4k.png" alt="MOSS-VL 量化配置 benchmark 对比" width="100%"/>
 </p>
 
 ## 硬件要求