Upload folder using huggingface_hub

README.md

@@ -7,7 +7,6 @@ tags:
 - hy3
 - moe
 - text-generation
-inference: false
 ---
 <p align="left">
     <a href="https://huggingface.co/tencent/Hy3/blob/main/README_CN.md">中文</a>&nbsp;｜&nbsp;English
@@ -42,8 +41,8 @@ inference: false
 ## Table of Contents
 
 - [Model Introduction](#model-introduction)
-- [Stronger Agent Performance](#stronger-agent-performance)
-- [Product Experience: More Reliable, More Cost-Effective](#product-experience-more-reliable-more-cost-effective)
+- [Stronger Agent Capabilities](#stronger-agent-capabilities)
+- [More Reliable Product Experiences](#more-reliable-product-experiences)
 - [Benchmark Appendix](#benchmark-appendix)
 - [News](#news)
 - [Model Links](#model-links)
@@ -60,7 +59,7 @@ inference: false
 
 ## Model Introduction
 
-**Hy3** is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team. Following the Hy3 Preview launch in late April, we gathered feedback from 50+ product teams. We fixed various issues in task execution and interaction, and improved both the quality and scale of our post-training pipeline. Today, we are launching Hy3. It significantly outperforms similar-size models and rivals flagship open-source models with 2-5x the parameters. It also shows solid gains in utility across productivity tasks and real-world applications.
+**Hy3** is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team. Following the Hy3 Preview launch in late April, we gathered feedback from 50+ products and scaled up post-training with higher quality data. Today, we introduce Hy3, which outperforms similar-size models and rivals flagship open-source models with 2-5x parameters. It also shows significant gains in utility across various products and productivity tasks.
 
 
 | Property | Value |
@@ -79,27 +78,27 @@ inference: false
 | Number of Experts | 192 experts, top-8 activated |
 | Supported Precisions | BF16 |
 
-## Stronger Agent Performance
+## Stronger Agent Capabilities
 
-Building on Hy3 Preview, we improved post-training data quality and diversity while scaling up RL training. Hy3 shows solid gains across reasoning, agentic workflows, and long-context tasks. Its performance is close to leading flagship models, both domestic and international.
+Building on Hy3 Preview, we further improved the quality and diversity of post-training data while scaling up RL training. Hy3 shows solid gains across reasoning, agentic, and long-context tasks, competitive with much larger flagship models.
 
 <p align="center">
   <img src="assets/benchmark.png" width="100%"/>
 </p>
 
-In productivity scenarios such as coding, document processing, financial analysis, game development, and frontend design, Hy3 has made solid gains, positioning it as a reliable, cost-effective option.
+In productivity scenarios such as coding, office work, financial modeling, frontend design, and game development, Hy3 has made remarkable progress and can now serve as a reliable, cost-effective model option.
 
-We don't think public benchmark scores tell the full story. So we ran a blind test with 270 experts from various disciplines, working on real-world workflows, and collected 312 valid comparisons. Hy3 scored 2.67/4, outperforming GLM-5.1 at 2.51/4. The advantage was clearest in frontend development, CI/CD, and data & storage.
+We don't think public benchmark scores tell the full story. So we ran a blind evaluation with 270 experts using tasks from their work, and Hy3 scored 2.67/4, outperforming GLM-5.1 at 2.51/4. The advantage was most substantial in frontend development, data & storage, and CI/CD tasks.
 
-## Product Experience: More Reliable, More Cost-Effective
+## More Reliable Product Experiences
 
-Utility in production is not fully captured by benchmarks. Based on extensive user feedback and product telemetry, we identified real-world behavior issues that break product experience and improved the model's capabilities in those areas, earning uniformly positive feedback from product teams.
+Model usefulness is not fully captured by benchmarks. Based on extensive product feedback, we identified and fixed the following issues, receiving consistently positive feedback from product teams.
 
-**Output Formatting and Tool Calling Stability**: We fixed multiple baseline reliability issues, bringing the model to production-grade standards across tool configurations and output constraints. Tool-call success rates and error recovery improved, and invalid calls that trigger infinite loops dropped. Hy3 also generalizes across different agent scaffoldings. On SWE-Bench Verified, accuracy variance across scaffoldings like CodeBuddy, Cline, and KiloCode remains within 4%.
+**Stability of tool calls and output formats**: We fixed multiple baseline reliability issues, bringing the model to production-grade standards across tool configurations and output constraints. Tool-call error recovery and overall efficiency improved. Hy3 also generalizes across different agent scaffoldings. On SWE-Bench Verified, accuracy variance across scaffoldings like CodeBuddy, Cline, and KiloCode remains within 4%.
 
-**World Knowledge and Anti-Hallucination**: Internal knowledge and external hallucination are interconnected and critical to real-world product experience. Guided by the ideal behavior pattern: "answer when grounded, state when evidence is missing, do not conflate sources, do not fabricate data," we implemented fine-grained data cleaning and specific training constraints. In internal evaluations on real-world scenarios, Hy3's hallucination rate dropped from 12.5% to 5.4%, and commonsense error rates fell from 25.4% to 12.7%. These improvements materially reduce fact conflation, fabrication, and logical contradiction.
+**Knowledge and anti-hallucination**: Guided by the ideal of "answer when grounded, state when evidence is missing, do not conflate sources or fabricate data," we implemented fine-grained data cleaning and training constraints. In internal evaluations based on real-world scenarios, Hy3's hallucination rate dropped from 12.5% to 5.4%, and commonsense error rates fell from 25.4% to 12.7%. These improvements materially reduce fact conflation, fabrication, and logical contradiction.
 
-**Complex Context Retention and Multi-turn Intent Tracking**: Through joint optimization of SFT and RL, Hy3 improved on operational pain points like coreference resolution, ellipsis recovery, and multi-turn constraint inheritance. On internal comprehensive multi-turn tests, the issue rate dropped from 17.4% to 7.9%. It also posted significant gains on open-source long-dialogue benchmarks like MRCR, from 42.9% to 75.1%. Overall outputs are more concise while ensuring complex intents do not decay or drift over long-horizon interactions.
+**Complex context retention and multi-turn intent tracking**: Through joint optimization of SFT and RL, Hy3 improved on operational pain points like coreference resolution, ellipsis recovery, and multi-turn constraint inheritance. On internal comprehensive multi-turn tests, the issue rate dropped from 17.4% to 7.9%. Hy3 also improved markedly on long-dialogue evals like MRCR. Its outputs are more concise while ensuring complex intents do not decay or drift over long-horizon interactions.
 
 ## Benchmark Appendix
 

README_CN.md

@@ -49,7 +49,7 @@
 
 ## 模型介绍
 
-**Hy3** 是由腾讯混元团队研发的快慢思考融合的混合专家模型，总参数量 295B，激活参数 21B，MTP 层参数 3.8B。4 月底发布 Hy3 Preview 后，我们在 50 多个业务中获得了广泛的反馈，修复了各种体验问题，进一步提升了后训练的质量和规模。今天，我们发布 Hy3。它展现出显著强于同尺寸并比肩旗舰（参数规模往往是 Hy3 的 2～5 倍）开源模型的智能水平，显著提升了在各类产品和生产力任务中的实用价值。
+**Hy3** 是由腾讯混元团队研发的快慢思考融合的混合专家模型，总参数量 295B，激活参数 21B，MTP 层参数 3.8B。4 月底发布 Hy3 Preview 后，我们在 50 多个业务中获得了广泛的反馈，修复了各种体验问题，进一步提升了后训练的质量和规模。今天，我们正式发布 Hy3。它展现出显著强于同尺寸模型的智能水平，并比肩更大尺寸旗舰模型的效果，大幅提升了在各类产品和生产力任务中的实用价值。
 
 
 | 属性 | 值 |
@@ -70,23 +70,23 @@
 
 ## 更强大的智能体能力
 
-Hy3 基于 Preview 进一步提升了后训练数据的质量和多样性，扩大了 RL 算力规模，在推理、智能体、长上下文等任务上显著进步，取得了比肩国内外更大尺寸旗舰模型的效果。
+Hy3 基于 Preview 进一步提升了后训练数据的质量和多样性，扩大了 RL 算力规模，在推理、智能体、长上下文等任务上显著进步，取得了比肩国内外更大尺寸旗舰模型（参数规模往往是 Hy3 的 2～5 倍）的效果。
 
 <p align="center">
   <img src="assets/benchmark.png" width="100%"/>
 </p>
 
-在代码、办公、金融、前端、游戏制作等生产力任务中，Hy3 进步显著，可以成为可靠且高性价比的模型选择。我们在内部组织了来自不同工种的 270 位专家在真实工作中盲测模型效果，产生了 312 条有效比较。结果表明，Hy3（均分 2.67 / 4）展现出优于 GLM5.1（均分 2.51 / 4）的表现，尤其在前端、CI/CD、数据与存储等类别优势显著。
+Hy3 在软件开发、办公生产、金融建模、前端设计、游戏制作等生产力任务上的进步尤其显著，可以成为高性价比的可靠选择。在内部组织的 270 位专家基于真实工作的模型盲测中，Hy3（均分 2.67 / 4）展现出优于 GLM5.1（均分 2.51 / 4）的表现，尤其在前端、数据与存储、CI/CD 等类别优势显著。
 
 ## 更可靠的产品体验
 
 模型的实用体验不完全与榜单成绩挂钩。基于广泛的用户反馈和分析，我们定位并优化了一系列体验向能力，获得了产品侧一致且积极的评价。
 
-**输出格式和工具调用稳定性**：我们显著改善了一系列基础底线问题，确保模型在各种工具设置和输出要求下达到生产级标准，工具调用的错误恢复能力和效率大幅提升。另外，Hy3 还增强了跨智能体环境的泛化性，跨框架泛化能力增强，不同脚手架（如 CodeBuddy、Cline、KiloCode）都能稳定发挥，在 SWE Bench Verified 上标准差整体控制在 4 个百分点以内。
+**输出格式和工具调用稳定性**：我们显著改善了一系列基础底线问题，确保模型在各种工具设置和输出要求下达到生产级标准，工具调用的错误恢复能力和效率大幅提升。另外，Hy3 还增强了跨脚手架泛化性，不同脚手架（如 CodeBuddy、Cline、KiloCode）在 SWE Bench Verified 上的分数标准差控制在 4 个百分点以内。
 
-**知识常识和抗幻觉能力**：内生知识和外部幻觉问题相互关联，且对于真实体验至关重要。基于“有依据才回答、无依据明示缺失，多来源信息不乱拼，数据和状态不乱编”的理想态，我们进行了细粒度的数据清洗和训练约束。在基于真实产品的内部评测中，Hy3 的幻觉率从 12.5% 降至 5.4%，常识错误率从 25.4% 降至 12.7%，显著改善了“张冠李戴”、无中生有、逻辑矛盾等问题。
+**知识常识和抗幻觉能力**：基于“有依据才回答、无依据明示缺失，多来源信息不乱拼，数据和状态不乱编”的理想态，我们进行了细粒度的数据清洗和训练约束。在基于真实产品的内部评测中，Hy3 的幻觉率从 12.5% 降至 5.4%，常识错误率从 25.4% 降至 12.7%，显著改善了“张冠李戴”、无中生有、逻辑矛盾等问题。
 
-**复杂上下文承接与多轮意图保持能力**：Hy3 在 SFT 与 RL 阶段联合优化了指代消解、省略还原及多轮约束继承等业务痛点问题，内部评测的多轮问题率从 17.4% 降至 7.9%。同时 Hy3 在长对话理解基准中取得显著跨越（如 MRCR 从 42.9% 升至 75.1%），整体输出更加精炼的同时确保复杂意图在长程交互中不衰减、不跑偏。
+**复杂上下文承接与多轮意图保持能力**：Hy3 在 SFT 与 RL 阶段联合优化了指代消解、省略还原���多轮约束继承等业务痛点问题，内部评测的多轮问题率从 17.4% 降至 7.9%。同时 Hy3 在长对话理解基准中取得显著跨越（如 MRCR 从 42.9% 升至 75.1%），输出更精炼的同时确保复杂意图在长程交互中不衰减、不跑偏。
 
 ## Benchmark 附录