Add Performance tools

.github/workflows/docker/docker-compose.yaml

@@ -1,6 +1,6 @@
 services:
   trinity-node-1:
-    image: trinity-rft-unittest:20260420
+    image: trinity-rft-unittest:20260506
     cap_add:
       - SYS_PTRACE
     pull_policy: never
@@ -34,7 +34,7 @@ services:
             capabilities: [gpu]
 
   trinity-node-2:
-    image: trinity-rft-unittest:20260420
+    image: trinity-rft-unittest:20260506
     cap_add:
       - SYS_PTRACE
     pull_policy: never

docs/sphinx_doc/source/tutorial/example_megatron.md

@@ -39,10 +39,10 @@ We provide a Docker setup to simplify environment management.
 #### Build the Docker Image
 
 
-Trinity-RFT provides a dedicated Dockerfile for Megatron-LM located at `scripts/docker/Dockerfile.megatron`. You can build the image using the following command:
+Trinity-RFT's provided Docker already has Megatron-LM related dependencies pre-installed. You can either use our provided Docker image directly or customize the Dockerfile to build your own image as needed.
 
 ```bash
-docker build -f scripts/docker/Dockerfile.megatron -t trinity-rft-megatron:latest .
+docker build -f scripts/docker/Dockerfile.uv -t trinity-rft-megatron:latest .
 ```
 
 > 💡 You can customize the Dockerfile before building — for example, to add pip mirrors or set API keys.
@@ -60,6 +60,7 @@ docker run -it \
 ```
 
 Replace `<your_data_and_checkpoints_path>` with the actual path on your machine where datasets and model checkpoints are stored.
+The image uses `uv` to manage Python dependencies, the virtual environment will be automatically activated after entering the docker container (you can also manually activate it with `source /opt/venv/bin/activate`). The image has include dependencies such as vllm, flash-attn and Megatron-LM, if you need to download more packages, don't forget to activate the virtual environment and use `uv pip install` to install them.
 
 ---
 

docs/sphinx_doc/source/tutorial/metrics_reference.md

@@ -164,7 +164,7 @@ This category includes metrics that track the training dynamics of the policy (a
 This category includes metrics that track the processing of experiences through various pipeline operators (`experience_pipeline/`) and data sampling statistics (`sample/`). These metrics are aggregated at the step level, as the experience pipeline and data sampling are performed in each step.
 
 
-#### Experience Pipeline Metrics (`experience_pipeline/` and `time/experience_pipeline/`)
+#### Experience Pipeline Metrics (`experience_pipeline/` and `experience_pipeline/time/`)
 
 Experience pipeline metrics track the processing of experiences through various pipeline operators. Each metric represents the count of the specific operator in one step.
 

docs/sphinx_doc/source/tutorial/trinity_installation.md

@@ -126,7 +126,7 @@ cd Trinity-RFT
 
 # Build the Docker image
 ## Tip: You can modify the Dockerfile to add mirrors or set API keys
-docker build -f scripts/docker/Dockerfile -t trinity-rft:latest .
+docker build -f scripts/docker/Dockerfile.uv -t trinity-rft:latest .
 
 # Run the container, replacing <path_to_your_data_and_checkpoints> with your actual path
 docker run -it \

docs/sphinx_doc/source_zh/tutorial/example_megatron.md

@@ -44,11 +44,10 @@ pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation \
 
 #### 构建 Docker 镜像
 
-Trinity-RFT 提供了专门用于 Megatron-LM 的 Dockerfile，位于 `scripts/docker/Dockerfile.megatron`。
-可以使用以下命令构建镜像：
+Trinity-RFT 提供的 Docker 已经预装了 Megatron-LM 相关依赖。你可以直接使用我们提供的 Docker 镜像，或者根据需要自定义 Dockerfile 来构建镜像。
 
 ```bash
-docker build -f scripts/docker/Dockerfile.megatron -t trinity-rft-megatron:latest .
+docker build -f scripts/docker/Dockerfile.uv -t trinity-rft:latest .
 ```
 
 > 💡 你可以在构建前自定义 Dockerfile —— 例如添加 pip 镜像源或设置 API 密钥。
@@ -62,10 +61,11 @@ docker run -it \
     --rm \
     -v $PWD:/workspace \
     -v <your_data_and_checkpoints_path>:/data \
-    trinity-rft-megatron:latest
+    trinity-rft:latest
 ```
 
 请将 `<your_data_and_checkpoints_path>` 替换为你机器上存储数据集和模型检查点的实际路径。
+该镜像使用 `uv` 来管理 Python 依赖，进入容器后虚拟环境会自动激活（也可通过 `source /opt/venv/bin/activate` 手动激活）。该镜像已经包含了 vllm, flash-attn 以及 Megatron-LM，如果需要使用其他依赖，可直接使用 `uv pip install` 来安装它们。
 
 ---
 

docs/sphinx_doc/source_zh/tutorial/metrics_reference.md

@@ -165,7 +165,7 @@ graph TD
 此类别包括跟踪通过各种数据处理操作（`experience_pipeline/`）和数据采样统计（`sample/`）的指标。这些指标在步骤（step）级别计算，因为 experience 处理和数据采样会在每个步骤中执行一次。
 
 
-#### Experience Pipeline 相关指标（`experience_pipeline/` 和 `time/experience_pipeline/`）
+#### Experience Pipeline 相关指标（`experience_pipeline/` 和 `experience_pipeline/time/`）
 
 Experience Pipeline 相关的指标统计了和数据处理相关的值，每个指标表示一个步骤中特定操作的统计值。
 

docs/sphinx_doc/source_zh/tutorial/trinity_installation.md

@@ -127,7 +127,7 @@ cd Trinity-RFT
 
 # 构建 Docker 镜像
 ## 提示：可根据需要修改 Dockerfile 添加镜像源或设置 API 密钥
-docker build -f scripts/docker/Dockerfile -t trinity-rft:latest .
+docker build -f scripts/docker/Dockerfile.uv -t trinity-rft:latest .
 
 # 运行容器，请将 <path_to_your_data_and_checkpoints> 替换为实际需要挂载的路径
 docker run -it \