diff --git a/README.md b/README.md
index d0d82f4..71b13d4 100644
--- a/README.md
+++ b/README.md
@@ -76,7 +76,7 @@ The training GPU implements environment for this README is as follows: models ar
 
 1. **Import the Dataset**
 
-   First, upload the dataset to the target folder `datasets`. After uploading, the folder structure (for example, under the `cifar10` folder, there are folders for each class; `class0` folder contains all images for class 0) should look like the following:
+   First, upload the dataset to the target folder `datasets` [**[issue]**](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1882902085). After uploading, the folder structure (for example, under the `cifar10` folder, there are folders for each class; `class0` folder contains all images for class 0) should look like the following:
 
    ```yaml
     datasets
@@ -145,7 +145,7 @@ The training GPU implements environment for this README is as follows: models ar
    python train.py --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
    ```
 4. Wait for the training to complete.
-5. If the training is interrupted due to any reason, you can resume it by setting `--resume` to `True` in the `train.py` file, specifying the epoch number where the interruption occurred, providing the folder name of the interrupted training (`run_name`), and running the file again. Alternatively, you can use the following command to resume the training:  
+5. If the training is interrupted due to any reason **[[issue]](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1882912391)**, you can resume it by setting `--resume` to `True` in the `train.py` file, specifying the epoch number where the interruption occurred, providing the folder name of the interrupted training (`run_name`), and running the file again. Alternatively, you can use the following command to resume the training:  
    **Conditional Resume Training Command**
 
    ```bash
@@ -168,7 +168,14 @@ The training GPU implements environment for this README is as follows: models ar
    # This is not using --start_epoch, default use last checkpoint 
    python train.py --resume True --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
    ```
-   
+6. The pretrained models are released with every major [Release](https://github.com/chairc/Integrated-Design-Diffusion-Model/releases), so please stay updated. To use a pretrained model [**[issue]**](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1886403967), download the model corresponding to parameters such as `network`, `image_size`, `act`, etc., and save it to any local folder. Adjust the `--pretrain` and `--pretrain_path` in the `train.py` file accordingly. You can also use the following command for training with a pretrained model:
+
+   **Command for training with a pretrained model**
+
+   ```bash
+   python train.py --pretrain True --pretrain_path /your/pretrain/path/model.pt --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
+   ```
+
 #### Distributed Training
 
 1. The basic configuration is similar to regular training, but note that enabling distributed training requires setting `--distributed` to `True`. To prevent arbitrary use of distributed training, we have several conditions for enabling distributed training, such as `args.distributed`, `torch.cuda.device_count() > 1`, and `torch.cuda.is_available()`.
@@ -212,7 +219,7 @@ The training GPU implements environment for this README is as follows: models ar
 | --amp                 |             | Automatic mixed precision training | bool | Enable automatic mixed precision training. It effectively reduces GPU memory usage but may affect training accuracy and results |
 | --optim                |             | Optimizer                       | str  | Optimizer selection. Currently supports Adam and AdamW       |
 | --act | | Activation function | str | Activation function selection. Currently supports gelu, silu, relu, relu6 and lrelu |
-| --lr                   |             | Learning rate                   | float | Initial learning rate. Currently only supports linear learning rate |
+| --lr                   |             | Learning rate                   | float | Initial learning rate. |
 | --lr_func | | Learning rate schedule | str | Setting learning rate schedule, currently supporting linear, cosine, and warmup_cosine. |
 | --result_path          |             | Save path                       | str  | Path to save the training results                            |
 | --save_model_interval  |             | Save model after each training  | bool | Whether to save the model after each training iteration for model selection based on visualization |
@@ -237,16 +244,28 @@ The training GPU implements environment for this README is as follows: models ar
 
 2. Set the necessary parameters such as `--conditional`, `--generate_name`, `--num_images`, `--num_classes`, `--class_name`, `--image_size`, `--result_path`, etc. If no parameters are set, the default settings will be used. There are two ways to set the parameters: one is to directly modify the `parser` in the `if __name__ == "__main__":` section of the `generate.py` file, and the other is to use the following commands in the console while in the `/your/path/Defect-Diffusion-Model/tools` directory:
 
-   **Conditional Generation Command**
+   **Conditional Generation Command (version > 1.1.1)**
+
+   ```bash
+   python generate.py --generate_name df --num_images 8 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt
+   ```
+
+   **Unconditional Generation Command (version > 1.1.1)**
+
+   ```bash
+   python generate.py --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt
+   ```
+
+   **Conditional Generation Command (version <= 1.1.1)**
 
    ```bash
-   python generate.py --conditional True --generate_name df --num_images 8 --num_classes 10 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt
+   python generate.py --conditional True --generate_name df --num_images 8 --num_classes 10 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt --sample ddpm --network unet --act gelu 
    ```
 
-   **Unconditional Generation Command**
+   **Unconditional Generation Command (version <= 1.1.1)**
 
    ```bash
-   python generate.py --conditional False --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt
+   python generate.py --conditional False --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt --sample ddpm --network unet --act gelu 
    ```
 
 3. Wait for the generation process to complete.
@@ -263,10 +282,10 @@ The training GPU implements environment for this README is as follows: models ar
 | --num_images    |             | Number of generated images      | int  | Number of images to generate                                 |
 | --weight_path   |             | Path to model weights           | str  | Path to the model weights file, required for network generation |
 | --result_path   |             | Save path                       | str  | Path to save the generated images                            |
-| --sample        |             | Sampling method                 | str  | Set the sampling method type, currently supporting DDPM and DDIM. |
-| --network       |             | Training network                | str  | Set the training network, currently supporting UNet, CSPDarkUNet. |
-| --act           |             | Activation function             | str  | Activation function selection. Currently supports gelu, silu, relu, relu6 and lrelu. If you do not set the same activation function as the model, mosaic phenomenon will occur. |
-| --num_classes   |      ✓      | Number of classes               | int  | Number of classes for classification                         |
+| --sample        |             | Sampling method                 | str  | Set the sampling method type, currently supporting DDPM and DDIM. **(No need to set for models after version 1.1.1)** |
+| --network       |             | Training network                | str  | Set the training network, currently supporting UNet, CSPDarkUNet. **(No need to set for models after version 1.1.1)** |
+| --act           |             | Activation function             | str  | Activation function selection. Currently supports gelu, silu, relu, relu6 and lrelu. If you do not set the same activation function as the model, mosaic phenomenon will occur. **(No need to set for models after version 1.1.1)** |
+| --num_classes   |      ✓      | Number of classes               | int  | Number of classes for classification **(No need to set for models after version 1.1.1)** |
 | --class_name    |      ✓      | Class name                      | int  | Index of the class to generate images. if class name is `-1`, the model would output one image per class. |
 | --cfg_scale     |      ✓      | Classifier-free guidance weight | int  | Weight for classifier-free guidance interpolation, for better generation model performance |
 
@@ -338,7 +357,7 @@ We conducted training on the following four datasets using the `DDPM` sampler wi
 
 #### Base on the 64×64 model to generate 160×160 (every size) images
 
-Of course, based on the 64×64 U-Net model, we generate 160×160 `NEU-DET` images in the `generate.py` file (single output, each image occupies 21GB of GPU memory). Detailed images are as follows:
+Of course, based on the 64×64 U-Net model, we generate 160×160 `NEU-DET` images in the `generate.py` file (single output, each image occupies 21GB of GPU memory). **Attention this [[issues]](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1886422210)**! If it's an image with defect textures where the features are not clear, generating a large size directly might not have these issues, such as in NRSD or NEU datasets. However, if the image contains a background with specific distinctive features, you may need to use super-resolution or resizing to increase the size, for example, in Cifar10, CelebA-HQ, etc. **If you really need large-sized images, you can directly train with large pixel images if there is enough GPU memory.** Detailed images are as follows:
 
 ![model_499_ema](assets/neu160_0.jpg)![model_499_ema](assets/neu160_1.jpg)![model_499_ema](assets/neu160_2.jpg)![model_499_ema](assets/neu160_3.jpg)![model_499_ema](assets/neu160_4.jpg)![model_499_ema](assets/neu160_5.jpg)
 
diff --git a/README_zh.md b/README_zh.md
index 76c1a4b..71395e1 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -76,7 +76,7 @@ Integrated Design Diffusion Model
 
 1. **导入数据集** 
 
-   首先，将数据集上传至目标文件夹`datasets`中。上传后文件夹格式（例如：cifar10文件夹下存放着所有类别；class0文件夹下存储着class0这个类别的所有图片）如下方列表所示：
+   首先，将数据集上传至目标文件夹`datasets`中[**[issue]**](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1882902085)。上传后文件夹格式（例如：cifar10文件夹下存放着所有类别；class0文件夹下存储着class0这个类别的所有图片）如下方列表所示：
    ```yaml
     datasets
     └── cifar10
@@ -144,7 +144,7 @@ Integrated Design Diffusion Model
    python train.py --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
    ```
 4. 等待训练即可
-5. 若因异常原因中断训练，我们可以在`train.py`文件，首先将`--resume`设置为`True`，其次设置异常中断的迭代编号，再写入该次训练的所在文件夹（run_name），最后运行文件即可。也可以使用如下命令进行恢复：
+5. 若因异常原因中断训练**[[issue]](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1882912391)**，我们可以在`train.py`文件，首先将`--resume`设置为`True`，其次设置异常中断的迭代编号，再写入该次训练的所在文件夹（run_name），最后运行文件即可。也可以使用如下命令进行恢复：
    **有条件恢复训练命令**
 
    ```bash
@@ -166,6 +166,12 @@ Integrated Design Diffusion Model
    # 此处为不输入--start_epoch参数，默认使用last权重
    python train.py --resume True --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
    ```
+6. 预训练模型在每次大版本[Release](https://github.com/chairc/Integrated-Design-Diffusion-Model/releases)中发布，请留意。预训练模型使用方法如下[**[issue]**](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1886403967)，首先将对应`network`、`image_size`、`act`等相同参数的模型下到本地任意文件夹下。直接调整`train.py`中`--pretrain`和`--pretrain_path`即可。也可以使用如下命令进行预训练：
+   **使用预训练模型训练命令**
+
+   ```bash
+   python train.py --pretrain True --pretrain_path /your/pretrain/path/model.pt --sample ddpm --conditional False --run_name df --epochs 300 --batch_size 16 --image_size 64 --dataset_path /your/dataset/path --result_path /your/save/path
+   ```
 
 #### 分布式训练
 
@@ -213,7 +219,7 @@ Integrated Design Diffusion Model
 | --amp                  |          | 混合精度训练                     |   bool   | 开启混合精度训练，有效减少显存使用，但无法保证训练精度和训练结果 |
 | --optim                |          | 优化器                           |   str    | 优化器选择，目前支持adam和adamw                              |
 | --act                  |          | 激活函数                         |   str    | 激活函数选择，目前支持gelu、silu、relu、relu6和lrelu         |
-| --lr                   |          | 学习率                           |  float   | 初始化学习率，目前仅支持线性学习率                           |
+| --lr                   |          | 学习率                           |  float   | 初始化学习率                                                 |
 | --lr_func              |          | 学习率方法                       |   str    | 设置学习率方法，当前支持linear、cosine和warmup_cosine        |
 | --result_path          |          | 保存路径                         |   str    | 保存路径                                                     |
 | --save_model_interval  |          | 是否每次训练储存                 |   bool   | 是否每次训练储存，根据可视化生成样本信息筛选模型             |
@@ -238,18 +244,29 @@ Integrated Design Diffusion Model
 1. 打开`generate.py`文件，找到`--weight_path`参数，将参数中的路径修改为模型权重路径，例如`/your/path/weight/model.pt`
 
 2. 设置必要参数，例如`--conditional`，`--generate_name`，`--num_images`，`--num_classes`，`--class_name`，`--image_size`，`--result_path`等参数，若不设置参数则使用默认设置。我们有两种参数设置方法，其一是直接对`generate.py`文件`if __name__ == "__main__":`中的`parser`进行设置；其二是在控制台在`/your/path/Defect-Diffiusion-Model/tools`路径下输入以下命令：
-   **有条件生成命令**
+   **有条件生成命令（1.1.1版本以上）**
 
    ```bash
-   python generate.py --conditional True --generate_name df --num_images 8 --num_classes 10 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt
+   python generate.py --generate_name df --num_images 8 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt
    ```
 
-   **无条件生成命令**
+   **无条件生成命令（1.1.1版本以上）**
 
    ```bash
-   python generate.py --conditional False --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt
+   python generate.py --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt
    ```
-
+   **有条件生成命令（1.1.1版本及以下）**
+   
+   ```bash
+   python generate.py --conditional True --generate_name df --num_images 8 --num_classes 10 --class_name 0 --image_size 64 --weight_path /your/path/weight/model.pt --sample ddpm --network unet --act gelu 
+   ```
+   
+   **无条件生成命令（1.1.1版本及以下）**
+   
+   ```bash
+   python generate.py --conditional False --generate_name df --num_images 8 --image_size 64 --weight_path /your/path/weight/model.pt --sample ddpm --network unet --act gelu 
+   ```
+   
 3. 等待生成即可
 
 #### 生成参数
@@ -264,10 +281,10 @@ Integrated Design Diffusion Model
 | --num_images    |          | 生成图片个数                     |   int    | 单次生成图片个数                                             |
 | --weight_path   |          | 权重路径                         |   str    | 模型权重路径，网络生成需要加载文件                           |
 | --result_path   |          | 保存路径                         |   str    | 保存路径                                                     |
-| --sample        |          | 采样方式                         |   str    | 设置采样器类别，当前支持ddpm，ddim                           |
-| --network       |          | 训练网络                         |   str    | 设置训练网络，当前支持UNet，CSPDarkUNet                      |
-| --act           |          | 激活函数                         |   str    | 激活函数选择，目前支持gelu、silu、relu、relu6和lrelu。如果不选择，会产生马赛克现象 |
-| --num_classes   |    是    | 类别个数                         |   int    | 类别个数，用于区分类别                                       |
+| --sample        |          | 采样方式                         |   str    | 设置采样器类别，当前支持ddpm，ddim**（1.1.1版本后的模型可不用设置）** |
+| --network       |          | 训练网络                         |   str    | 设置训练网络，当前支持UNet，CSPDarkUNet**（1.1.1版本后的模型可不用设置）** |
+| --act           |          | 激活函数                         |   str    | 激活函数选择，目前支持gelu、silu、relu、relu6和lrelu。如果不选择，会产生马赛克现象**（1.1.1版本后的模型可不用设置）** |
+| --num_classes   |    是    | 类别个数                         |   int    | 类别个数，用于区分类别**（1.1.1版本后的模型可不用设置）**    |
 | --class_name    |    是    | 类别名称                         |   int    | 类别序号，用于对指定类别生成。如果输入为-1，则模型为每类输出一张图片 |
 | --cfg_scale     |    是    | classifier-free guidance插值权重 |   int    | classifier-free guidance插值权重，用户更好生成模型效果       |
 
@@ -341,7 +358,7 @@ Integrated Design Diffusion Model
 
 #### 基于64×64模型生成160×160（任意大尺寸）图像
 
-当然，我们根据64×64的基础模型，在`generate.py`文件中生成160×160的`NEU-DET`图片（单张输出，每张图片占用显存21GB）。详细图片如下：
+当然，我们根据64×64的基础模型，在`generate.py`文件中生成160×160的`NEU-DET`图片（单张输出，每张图片占用显存21GB）。**请注意这个**[[**issue**]](https://github.com/chairc/Integrated-Design-Diffusion-Model/issues/9#issuecomment-1886422210)：如果是缺陷纹理那种图片，特征物不明显的直接生成大尺寸就不会有这些问题，例如NRSD、NEU数据集。如果是含有背景有特定明显特征的则需要超分或者resize提升尺寸，例如Cifar10、CelebA-HQ等。**如果实在需要大尺寸图像，在显存足够的情况下直接训练大像素图片。**详细图片如下：
 
 ![model_499_ema](assets/neu160_0.jpg)![model_499_ema](assets/neu160_1.jpg)![model_499_ema](assets/neu160_2.jpg)![model_499_ema](assets/neu160_3.jpg)![model_499_ema](assets/neu160_4.jpg)![model_499_ema](assets/neu160_5.jpg)
 
diff --git a/tools/generate.py b/tools/generate.py
index 4c70e21..164ae61 100644
--- a/tools/generate.py
+++ b/tools/generate.py
@@ -96,7 +96,8 @@ def generate(args):
     # Generation name (required)
     parser.add_argument("--generate_name", type=str, default="df")
     # Input image size (required)
-    # [Warn] Compatible with older versions, version <= 1.1.1 need to be equal to model's image size
+    # [Warn] Compatible with older versions
+    # [Warn] Version <= 1.1.1 need to be equal to model's image size, version > 1.1.1 can set whatever you want
     parser.add_argument("--image_size", type=int, default=64)
     # Number of generation images (required)
     # if class name is `-1` and conditional `is` True, the model would output one image per class.
@@ -112,18 +113,21 @@ def generate(args):
     # Set network
     # Option: unet/cspdarkunet
     # [Warn] Compatible with older versions, version <= 1.1.1
+    # [Warn] Version <= 1.1.1 need to be equal to model's network, version > 1.1.1 can set whatever you want
     parser.add_argument("--network", type=str, default="unet")
     # Set activation function (needed)
     # [Note] The activation function settings are consistent with the loaded model training settings.
     # [Note] If you do not set the same activation function as the model, mosaic phenomenon will occur.
     # Option: gelu/silu/relu/relu6/lrelu
     # [Warn] Compatible with older versions, version <= 1.1.1
+    # [Warn] Version <= 1.1.1 need to be equal to model's act, version > 1.1.1 can set whatever you want
     parser.add_argument("--act", type=str, default="gelu")
 
     # =====================Enable the conditional generation (if '--conditional' is set to 'True')=====================
     # Number of classes (required)
     # [Note] The classes settings are consistent with the loaded model training settings.
     # [Warn] Compatible with older versions, version <= 1.1.1
+    # [Warn] Version <= 1.1.1 need to be equal to model's num classes, version > 1.1.1 can set whatever you want
     parser.add_argument("--num_classes", type=int, default=10)
     # Class name (required)
     # if class name is `-1`, the model would output one image per class.