[doc] update README and doc.

Author: wangzhaode

README.md

@@ -11,6 +11,7 @@ It focuses on supporting the subset of Jinja2 used by modern Large Language Mode
 ## Features
 
 - **C++11 Compatible**: Ensures maximum compatibility across older compiler versions and embedded systems.
+- **Easy Integration**: The core library consists of just **one header file** (`include/jinja.hpp`) and **one source file** (`src/jinja.cpp`), making it extremely easy to copy and drop into any project.
 - **Lightweight**: Minimal dependencies (only `nlohmann/json`).
 - **LLM Focused**: Native support for `messages`, `tools`, `add_generation_prompt`, and special tokens.
 - **Strictly Typed**: Uses `nlohmann::json` for context management.
@@ -20,12 +21,13 @@ It focuses on supporting the subset of Jinja2 used by modern Large Language Mode
 ## Supported Models
 
 Tested and verified with templates from:
-- **Llama 3 / 3.1 / 3.2** (Instruct & Vision)
-- **Qwen 2.5** (Coder, Math, VL, Omni)
-- **Qwen 3** (Instruct, Thinking, QwQ)
+- **Qwen 2.5 / 3** (Coder, Math, VL, Omni, Instruct, Thinking, QwQ)
 - **DeepSeek** (V3, R1)
+- **Llama 3 / 3.1 / 3.2** (Instruct & Vision)
 - **Mistral**
 - **Gemma**
+- **SmolLM**
+- **Phi**
 - And more...
 
 ## Build Instructions

README_CN.md

@@ -11,6 +11,7 @@
 ## 特性
 
 - **C++11 兼容**：确保在旧版编译器和嵌入式系统上的最大兼容性。
+- **易于集成**：核心库仅包含**一个头文件** (`include/jinja.hpp`) 和**一个源文件** (`src/jinja.cpp`)，非常方便拷贝并集成到任何项目中。
 - **轻量级**：依赖极少 (仅依赖 `nlohmann/json`，已包含在项目中)。
 - **专注 LLM**：原生支持 `messages`, `tools`, `add_generation_prompt` 以及特殊 token 的处理。
 - **类型安全**：使用 `nlohmann::json` 进行上下文管理。
@@ -20,12 +21,13 @@
 ## 支持的模型
 
 已基于以下模型的真实模板进行测试验证：
-- **Llama 3 / 3.1 / 3.2** (Instruct & Vision)
-- **Qwen 2.5** (Coder, Math, VL, Omni)
-- **Qwen 3** (Instruct, Thinking, QwQ)
+- **Qwen 2.5 / 3** (Coder, Math, VL, Omni, Instruct, Thinking, QwQ)
 - **DeepSeek** (V3, R1)
+- **Llama 3 / 3.1 / 3.2** (Instruct & Vision)
 - **Mistral**
 - **Gemma**
+- **SmolLM**
+- **Phi**
 - 更多...
 
 ## 构建指南

doc/implementation_details.md

@@ -89,5 +89,20 @@ To support a wide range of deployment environments:
 
 ## Testing Strategy
 
-*   **Real Data**: We use `tests/test_chat_template.json` generated from the official Python `transformers` library on typically supported models.
-*   **Fuzzy Matching**: For dynamic content (like dates), tests use regex normalization to ensure pass consistency across time and environments.
+*   **Real Data**: We use `tests/test_chat_template.json` generated from the official Python `transformers` library.
+*   **Fuzzy Matching**: For dynamic content (like dates), tests use regex normalization to ensure pass consistency.
+
+### Tested Models
+The following models are automatically verified in our test suite:
+
+*   **Qwen**: `Qwen2.5-3B-Instruct`, `Qwen2.5-VL-3B-Instruct`, `Qwen2.5-Omni-3B`, `Qwen2.5-7B-Instruct-1M`, `Qwen2.5-Math-7B-Instruct`, `QwQ-32B`
+*   **Qwen3**: `Qwen3-4B`, `Qwen3-4B-Instruct`, `Qwen3-4B-Thinking`, `Qwen3-VL-4B-Instruct`, `Qwen3-VL-4B-Thinking`, `Qwen3Guard-Gen-4B`, `Qwen3-Coder-30B-A3B-Instruct`, `Qwen3-Omni-30B-A3B-Instruct`, `Qwen3-Omni-30B-A3B-Thinking`
+*   **DeepSeek**: `DeepSeek-R1-Distill-Qwen-7B`, `DeepSeek-V3.2`, `DeepSeek-R1`
+*   **GLM**: `ZhipuAI/GLM-4.5V`, `ZhipuAI/GLM-4.6V`
+*   **Yi**: `01ai/Yi-VL-6B`, `01ai/Yi-1.5-6B-Chat`
+*   **SmolLM**: `HuggingFaceTB/SmolLM-135M-Instruct`, `HuggingFaceTB/SmolVLM-256M-Instruct`, `HuggingFaceTB/SmolLM2-135M-Instruct`, `HuggingFaceTB/SmolLM3-3B`
+*   **Gemma**: `google/gemma-3-4b-it`, `google/gemma-3n-E4B-it`
+*   **Mistral**: `mistralai/Ministral-3-3B-Instruct-2512`
+*   **Llama**: `llama-2-7b`, `Meta-Llama-3-8B-Instruct`, `Llama-3.2-3B-Instruct`
+*   **Phi**: `Phi-3.5-mini-instruct`, `Phi-3.5-vision-instruct`, `phi-4`, `Phi-4-mini-reasoning`
+*   **MobileLLM**: `LLM-Research/MobileLLM-125M`

doc/implementation_details_CN.md

@@ -91,3 +91,18 @@ Lexer 部分模拟了 Jinja2 的 `lstrip_blocks` 和 `trim_blocks` 行为。手
 
 *   **真实数据**: 我们使用 `tests/test_chat_template.json`，该文件是从官方 Python `transformers` 库针对通常支持的模型生成的。
 *   **模糊匹配**: 对于动态内容 (如日期)，测试使用正则归一化来确跨时间和环境的通过一致性。
+
+### 已通过测试的模型 (Tested Models)
+我们在测试套件中自动验证了以下模型：
+
+*   **Qwen**: `Qwen2.5-3B-Instruct`, `Qwen2.5-VL-3B-Instruct`, `Qwen2.5-Omni-3B`, `Qwen2.5-7B-Instruct-1M`, `Qwen2.5-Math-7B-Instruct`, `QwQ-32B`
+*   **Qwen3**: `Qwen3-4B`, `Qwen3-4B-Instruct`, `Qwen3-4B-Thinking`, `Qwen3-VL-4B-Instruct`, `Qwen3-VL-4B-Thinking`, `Qwen3Guard-Gen-4B`, `Qwen3-Coder-30B-A3B-Instruct`, `Qwen3-Omni-30B-A3B-Instruct`, `Qwen3-Omni-30B-A3B-Thinking`
+*   **DeepSeek**: `DeepSeek-R1-Distill-Qwen-7B`, `DeepSeek-V3.2`, `DeepSeek-R1`
+*   **GLM**: `ZhipuAI/GLM-4.5V`, `ZhipuAI/GLM-4.6V`
+*   **Yi**: `01ai/Yi-VL-6B`, `01ai/Yi-1.5-6B-Chat`
+*   **SmolLM**: `HuggingFaceTB/SmolLM-135M-Instruct`, `HuggingFaceTB/SmolVLM-256M-Instruct`, `HuggingFaceTB/SmolLM2-135M-Instruct`, `HuggingFaceTB/SmolLM3-3B`
+*   **Gemma**: `google/gemma-3-4b-it`, `google/gemma-3n-E4B-it`
+*   **Mistral**: `mistralai/Ministral-3-3B-Instruct-2512`
+*   **Llama**: `llama-2-7b`, `Meta-Llama-3-8B-Instruct`, `Llama-3.2-3B-Instruct`
+*   **Phi**: `Phi-3.5-mini-instruct`, `Phi-3.5-vision-instruct`, `phi-4`, `Phi-4-mini-reasoning`
+*   **MobileLLM**: `LLM-Research/MobileLLM-125M`