* update docs

Author: lxowalle

docs/doc/en/mllm/dlm_lora_sdv1_5.md

@@ -23,7 +23,9 @@ LCM-LoRA-SDv1-5 is a model that supports text-to-image and image-to-image genera
 
 ### Model and Download Link
 
-If the `LCM-LoRA-SDv1-5` model is not present in the system directory `/root/models` by default, you need to download it manually.
+MaixPy currently supports `LCM-LoRA-SDv1-5`. Due to the large size of the model, you need to download the model manually and save it to the `/root/models` directory.
+
+> !!! Important !!! Important !!! The model must be saved in the `/root/models` directory; otherwise, some applications may fail to load it. For example, a valid save path is `/root/models/lcm-lora-sdv1-5-maixcam2`.
 
  * Memory requirement: CMM memory 1 GiB. For details, refer to the [Memory Usage Documentation](../pro/memory.md)
 

docs/doc/en/mllm/llm_deepseek.md

@@ -30,7 +30,9 @@ As mentioned above, since the network structure is the same as Qwen2.5, please r
 
 ### Models and Download Links
 
-By default, if `/root/models` directory in the system have no model, you can download it manually.
+MaixPy currently supports `deepseek-r1-distill-qwen-1.5B-maixcam2`. Due to the large size of the model, you need to download the model manually and save it to the `/root/models` directory.
+
+> !!! Important !!! Important !!! The model must be saved in the `/root/models` directory; otherwise, some applications may fail to load it. For example, a valid save path is `/root/models/deepseek-r1-distill-qwen-1.5B-maixcam2`.
 
 * **1.5B**:
 

docs/doc/en/mllm/vlm_qwen3.md

@@ -25,8 +25,9 @@ update:
 
 ### Model and Download Address
 
-MaixPy currently supports `Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448`. Due to the large model size, you need to download the model yourself and save it to the `/root/models directory`.
-> !!! IMPORTANT !!! IMPORTANT !!! The model MUST be saved in the `/root/models` directory, otherwise the model cannot be loaded. For example, the save path should be `/root/models/sipeed/Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2`
+MaixPy currently supports `Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448`. Due to the large size of the model, you need to download the model manually and save it to the `/root/models` directory.
+
+> !!! Important !!! Important !!! The model must be saved in the `/root/models` directory; otherwise, some applications may fail to load it. For example, a valid save path is `/root/models/Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2`.
 
 * **2B**:
   * Memory Requirement: 2GiB CMM Memory. Please refer to the Memory Usage Documentation for memory explanation.

docs/doc/zh/mllm/dlm_lora_sdv1_5.md

@@ -23,7 +23,8 @@ LCM-LoRA-SDv1-5 是一个支持文生图, 图生图的模型, 基于 StableDiffu
 
 ### 模型和下载地址
 
-默认系统`/root/models`目录下如果没有`LCM-LoRA-SDv1-5`模型，需要自行下载。
+MaixPy 目前支持 `LCM-LoRA-SDv1-5`，由于模型较大, 需要自行下载模型并保存到`/root/models`目录下。
+> !!! 注意 !!! 注意 !!! 模型一定要保存到`/root/models`目录下，否则一些应用可能无法加载模型. 例如保存路径为`/root/models/lcm-lora-sdv1-5-maixcam2`
 
   * 内存需求：CMM 内存 1GiB，内存解释请看[内存使用文档](../pro/memory.md)
   * 下载地址：https://huggingface.co/sipeed/lcm-lora-sdv1-5-maixcam2

docs/doc/zh/mllm/llm_deepseek.md

@@ -31,7 +31,8 @@ DeepSeek-R1 是深度科技（DeepSeek-AI）研发的大语言模型（LLM），
 
 ### 模型和下载地址
 
-默认系统`/root/models`目录下如果没有模型，可以自行下载。
+MaixPy 目前支持 `deepseek-r1-distill-qwen-1.5B`，由于模型较大, 需要自行下载模型并保存到`/root/models`目录下。
+> !!! 注意 !!! 注意 !!! 模型一定要保存到`/root/models`目录下，否则一些应用可能无法加载模型. 例如保存路径为`/root/models/deepseek-r1-distill-qwen-1.5B-maixcam2`
 
 * **1.5B**:
   * 内存需求：CMM 内存 1.8GiB，内存解释请看[内存使用文档](../pro/memory.md)

docs/doc/zh/mllm/vlm_qwen3.md

@@ -26,7 +26,8 @@ MaixPy 中移植了 [Qwen3-VL-2B](https://huggingface.co/Qwen/Qwen3-VL-2B-Instru
 ### 模型和下载地址
 
 MaixPy 目前支持 `Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448`，由于模型较大, 需要自行下载模型并保存到`/root/models`目录下。
-> !!! 注意 !!! 注意 !!! 模型一定要保存到`/root/models`目录下，否则无法加载模型. 例如保存路径为`/root/models/sipeed/Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2`
+> !!! 注意 !!! 注意 !!! 模型一定要保存到`/root/models`目录下，否则一些应用可能无法加载模型. 例如保存路径为`/root/models/Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2`
+
 * **2B**:
   * 内存需求： CMM 内存 2GiB，内存解释请看[内存使用文档](../pro/memory.md)
   * 下载地址： [Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2](https://huggingface.co/sipeed/Qwen3-VL-2B-Instruct-GPTQ-Int4-AX630C-P320-CTX448-maixcam2)