flagos-ai · zhzhcookie · Oct 20, 2025 · Oct 20, 2025 · Oct 20, 2025 · Oct 20, 2025
diff --git a/reports/v0.2.0/release_notes_v0.2.0.md b/reports/v0.2.0/release_notes_v0.2.0.md
@@ -1,14 +1,14 @@
-[中文版](./release_notes_v0.2.0_cn.md)
+<div align="right"><a href="./release_notes_v0.2.0_cn.md">中文版</a></div>
 
-## FlagTree 0.1.0 Release
+## FlagTree 0.2.0 Release
 
 ### Highlights
 
 FlagTree inherits capabilities from the previous version, continuously integrates new backends, expands support for Triton versions, and provides hardware-aware optimization capabilities. The project is currently in its early stages, aiming to be compatible with existing adaptation solutions for various AI chip backends, unify the code repository, build a code co-construction platform, and quickly implement multi-backend support in a single repository.
 
 ### New features
 
-* Added multi-Backend Support
+* Added multi-backend Support
 
 Currently supported backends include triton_shared cpu, iluvatar, xpu (klx), mthreads, __metax__, __aipu__(arm npu), __ascend__ npu & cpu, __tsingmicro__, cambricon, with __bold__ indicating newly added ones. <br>
 Each new backend maintains the capabilities of the previous version: cross-platform compilation and rapid verification, plugin-based high-differentiation modules, CI/CD, and quality management capabilities. <br>

diff --git a/reports/v0.2.0/release_notes_v0.2.0_cn.md b/reports/v0.2.0/release_notes_v0.2.0_cn.md
@@ -1,4 +1,4 @@
-[English](./release_notes_v0.2.0.md)
+<div align="right"><a href="./release_notes_v0.2.0.md">English</a></div>
 
 ## FlagTree 0.2.0 Release
 

diff --git a/reports/v0.3.0/release_notes_v0.3.0.md b/reports/v0.3.0/release_notes_v0.3.0.md
@@ -0,0 +1,36 @@
+<div align="right"><a href="./release_notes_v0.3.0_cn.md">中文版</a></div>
+
+## FlagTree 0.3.0 Release
+
+### Highlights
+
+FlagTree inherits capabilities from the previous version, continuously integrates new backends, and strengthens the ecosystem matrix. The project is currently in its early stage, aiming to be compatible with existing adaptation solutions for various chip backends, unify code repositories, create a collaborative code-building platform, and quickly achieve single-repository multi-backend support. Meanwhile, it continues to develop unified programming interface extensions, build intermediate layer representation and conversion extensions (FLIR), and enhance hardware awareness and compilation guidance support capabilities and scope (flagtree_hints).
+
+### New features
+
+* Added multi-backend Support
+
+Currently supported backends include triton_shared cpu, iluvatar, xpu (klx), mthreads, metax, aipu(arm npu), ascend npu & cpu, tsingmicro, cambricon, __hcu__, with __bold__ indicating newly added ones. <br>
+Each new backend maintains the capabilities of the previous version: cross-platform compilation and rapid verification, plugin-based high-differentiation modules, CI/CD, and quality management capabilities. <br>
+
+* Continuous integration with upstream ecosystems
+
+Thanks to the technical support from our partners, FlagTree has added compatibility with Paddle framework, OpenAnolis operating system, and Beijing Super Cloud Computing Center.
+
+* Continuous development of FLIR
+
+Ongoing expansion of DSL, TTIR extensions, Linalg intermediate representation and transformation extensions, and MLIR extensions to provide programming flexibility, enrich expression capabilities, and improve transformation capabilities.
+
+* Established compilation guidance specifications, added unified management module for multi-backend compilation
+
+flagtree_hints provides guidance for hardware unit mapping and compilation transformation optimization choices, and manages backend guidance differences through a unified module.
+
+* Joint construction with FlagGems operator library
+
+Collaborating with [FlagGems](https://github.com/FlagOpen/FlagGems) operator library on version compatibility, backend interfaces, registration mechanisms, and test modifications to support related features.
+
+### Looking ahead
+
+Improving GPGPU backend integration, decoupling backend specialization from main code implementation to establish an engineering foundation for FlagTree's general extensions and optimizations. <br>
+Aiming to comprehensively cover various implementation styles in the operator library, enhancing FLIR compilation completeness to match multiple backend requirements and enable compilation for more backends. <br>
+flagtree_hints will continue to explore operator performance optimization potential on different backends along both TritonGPU and Linalg compile-paths. <br>
diff --git a/reports/v0.3.0/release_notes_v0.3.0_cn.md b/reports/v0.3.0/release_notes_v0.3.0_cn.md
@@ -0,0 +1,36 @@
+<div align="right"><a href="./release_notes_v0.3.0.md">English</a></div>
+
+## FlagTree 0.3.0 Release
+
+### Highlights
+
+FlagTree 继承前一版本的能力，持续集成新的后端，壮大生态矩阵。项目当前处于初期，目标是兼容各芯片后端现有适配方案，统一代码仓库，打造代码共建平台，快速实现单仓库多后端支持。同时持续建设统一编程接口扩展，持续建设中间层表示及转换扩展（FLIR），提升硬件感知和编译指导支持能力与范围（flagtree_hints）。
+
+### New features
+
+* 新增多后端支持
+
+目前支持的后端包括 triton_shared cpu、iluvatar、xpu (klx)、mthreads、metax、aipu(arm npu)、ascend npu & cpu、tsingmicro、cambricon、__hcu__，其中 __加粗__ 为本次新增。 <br>
+各新增后端保持前一版本的能力：跨平台编译与快速验证、高差异度模块插件化、CI/CD、质量管理能力。 <br>
+
+* 持续对接上游生态
+
+感谢合作单位的技术支持，FlagTree 新增适配框架 Paddle、操作系统 OpenAnolis、北京超级云计算中心。
+
+* 持续建设 FLIR
+
+持续进行 DSL 扩展、TTIR 扩展、Linalg 中间层表示及转换扩展、MLIR 扩展，提供编程灵活性，丰富表达能力，完善转换能力。
+
+* 确定编译指导规范，新增多后端编译统一管理模块
+
+flagtree_hints 对硬件单元映射、编译转换优化选择进行指导，并通过统一模块管理多后端的指导差异。
+
+* 与 FlagGems 算子库联合建设
+
+在版本适配、后端接口、注册机制、测试修改等方面，与 [FlagGems](https://github.com/FlagOpen/FlagGems) 算子库联合支持相关特性。
+
+### Looking ahead
+
+GPGPU 后端完善接入整合，后端特化与主代码实现解耦，为应用 FlagTree 通用扩展、通用优化打下工程基础。 <br>
+以全面覆盖算子库中的多种写法为目标，完善 FLIR 编译的完备度，匹配多种后端需求，打通更多后端的编译。 <br>
+flagtree_hints 在 TritonGPU、Linalg 两条路线的不同后端上继续挖掘算子的性能优化潜力。 <br>
diff --git a/reports/v0.3.0/report_tests.md b/reports/v0.3.0/report_tests.md
@@ -0,0 +1,14 @@
+
+
+## FlagTree Test-Report
+
+FlagTree tests are validated on different backends, but currently the tests consist of only unit tests, which we will refine in the future for smaller or larger scale tests.
+
+### 1. Python unit test:
+
+| 　　　　　　　　　　　　 | default                   | iluvatar                                 | klx xpu                                       | mthreads                                       | metax                                       | hcu                                       |
+|----------------------|---------------------------|-------------------------------------------|------------------------------------------------|------------------------------------------------|---------------------------------------------|---------------------------------------------|
+| Number of unit tests | 9161 items               | 11395 items                               | 4183 items                                    | 4116 items                                    | 6309 items                                 | 309 items                                 |
+| Script location      | flagtree/python/test/unit | flagtree/third_party/iluvatar/python/test/unit | flagtree/third_party/xpu/python/test/unit | flagtree/third_party/mthreads/python/test/unit | flagtree/third_party/metax/python/test/unit | flagtree/third_party/hcu/python/test/unit |
+| Test command         | python3 -m pytest -s      | python3 -m pytest -s                      | python3 -m pytest -s                           | python3 -m pytest -s                           | python3 -m pytest -s                        | sh flagtree_test.sh                        |
+| Passing rate         | 100%                      | 100%                                      | 100%                                           | 100%                                           | 100%                                        | 100%                                        |