Merge pull request FlagAI-Open#490 from Anhforth/update_docs

docs update for new model version

examples/Aquila/Aquila-chat/README.md

@@ -26,7 +26,7 @@
 | AquilaCode-7B-NV          | 基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在英伟达芯片完成训练  |   AquilaCode-7B 以小数据集、小参数量，实现高性能，是目前支持中英双语的、性能最好的开源代码模型，经过了高质量过滤、使用有合规开源许可的训练代码数据进行训练。<br><br> AquilaCode-7B 分别在英伟达和国产芯片上完成了代码模型的训练。  | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  |[下载AquilaCode-7B-NV](https://model.baai.ac.cn/model-detail/100102)  | 已发布  | Nvidia-A100 | 
 | AquilaCode-7B-TS           |基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在天数智芯芯片上完成训练  |    同上    | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  | [下载AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | 已发布  | Tianshu-BI-V100  | 
 
-悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的 `model_pytorch.bin`，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)** 。
+悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的`checkpoints_in/aquilachat-7b`，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)**
 
 <br>如有使用问题请先查看 [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371)，若不能解决，请直接提交 [issue](https://github.com/FlagAI-Open/FlagAI/issues) ~
 
@@ -131,7 +131,6 @@ python generate_chat_bminf.py
     aquila-7b                模型名称，注意需要小写
     aquila_experiment        实验名称，可自定义
     ```
-    **如果启动deepspeed微调(在单张V100上运行微调为例)，上一步改为运行**
 
     **如果启动LoRA微调(在单张V100上运行微调为例)，上一步改为运行**
     ```
@@ -166,7 +165,7 @@ python generate_chat_bminf.py
 | warm_up           |float |   初始学习率与原始学习率的比例;     | 
 | save_interval         | int  |   模型保存的间隔，即每训练多少个iteration保存一次模型。当训练时间较长时，保存间隔可以避免因突然中断或出现错误导致训练成果全部丢失;   | 
 | log_interval           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
-| lora           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
+| lora           |bool  |    是否启用lora微调   | 
 | enable_sft_dataset_dir           |str  |    SFT训练数据集的目录    | 
 | enable_sft_dataset_file           |str  |    SFT训练数据集的文件名    | 
 

examples/Aquila/Aquila-chat/README_en.md

@@ -28,7 +28,7 @@ The additional details of the Aquila model will be presented in the official tec
 | AquilaCode-7B-TS   | Base model, "text-code" generation model, further pre-trained based on Aquila-7B, trained on Horizon Robotics chips | Same as above | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code) | [Download AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | Released        | Tianshu-BI-V100 |
 
 
-We will continue to release improved versions of Aquila model as open source. You can start by deleting the `model_pytorch.bi`n file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
+We will continue to release improved versions of Aquila model as open source. You can start by deleting the model checkpoint file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
 
 
  <br>If you have any question, please refer to the [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371) first. If you cannot solve them, please submit an [issue](https://github.com/FlagAI-Open/FlagAI/issues) directly.
@@ -130,7 +130,7 @@ Note: The Aquila-7B basic model may not perform as well for dialogue reasoning t
     ```
     bash dist_trigger_docker.sh hostfile Aquila-chat-lora.yaml aquila-7b aquila_experiment
     ```
-    The model trained using LoRa needs to be inferred using "generate_chat_lora.py", and the Lora parameters used during training should be added when loading the model in the autoloader.
+    Note: When training Lora, it will generate an `adapter_config.json` and `adapter_model.bin` file, located in the output directory (at the same level as the log file). For inference, please run the `Aquila-chat/generate_chat_lora.py` file. The difference compared to regular inference is that the autoloader, when loading the model for inference, requires specifying the directory of the adapter files in the `adapter_dir` parameter.
 <details><summary>The correct output information is shown below:</summary>
 
 The following information will be output. Note that `NODES_NUM` should be equal to the number of nodes, and `LOGFILE` is the log file for the model run.
@@ -159,7 +159,7 @@ For the above examples, you can modify the following parameters to achieve diffe
 | warm_up        | float | The ratio of the initial learning rate to the original learning rate. |
 | save_interval  | int   | The interval at which the model is saved, that is, how often the model is saved every few iterations of training. When the training time is long, the save interval can prevent all training results from being lost due to sudden interruptions or errors. |
 | log_interval   | int   | The interval at which logs are output, that is, how often log information is output every few iterations of training. |
-| lora           | int   | An integer value to enable LoRA optimization method during training. By default, it is set to 0 (no LoRA).|
+| lora           |bool   | Whether to enable LoRA optimization method during training. By default, it is set to False (no LoRA).|
 | enable_sft_dataset_dir | str | The directory of the SFT training dataset. |
 | enable_sft_dataset_file | str | The file name of the SFT training dataset. 
 

examples/Aquila/Aquila-code/README.md

@@ -26,7 +26,7 @@
 | AquilaCode-7B-NV          | 基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在英伟达芯片完成训练  |   AquilaCode-7B 以小数据集、小参数量，实现高性能，是目前支持中英双语的、性能最好的开源代码模型，经过了高质量过滤、使用有合规开源许可的训练代码数据进行训练。<br><br> AquilaCode-7B 分别在英伟达和国产芯片上完成了代码模型的训练。  | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  |[下载AquilaCode-7B-NV](https://model.baai.ac.cn/model-detail/100102)  | 已发布  | Nvidia-A100 | 
 | AquilaCode-7B-TS           |基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在天数智芯芯片上完成训练  |    同上    | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  | [下载AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | 已发布  | Tianshu-BI-V100  | 
 
-悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的 `model_pytorch.bin`，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)** 。
+悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的模型，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)** 。
 
 <br>如有使用问题请先查看 [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371)，若不能解决，请直接提交 [issue](https://github.com/FlagAI-Open/FlagAI/issues) ~
 
@@ -154,7 +154,7 @@ python generate_code_bminf.py
 | warm_up           |float |   初始学习率与原始学习率的比例;     | 
 | save_interval         | int  |   模型保存的间隔，即每训练多少个iteration保存一次模型。当训练时间较长时，保存间隔可以避免因突然中断或出现错误导致训练成果全部丢失;   | 
 | log_interval           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
-| lora           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
+| lora           |bool  |    是否启用lora微调   | 
 | enable_sft_dataset_dir           |str  |    SFT训练数据集的目录    | 
 | enable_sft_dataset_file           |str  |    SFT训练数据集的文件名    | 
 

examples/Aquila/Aquila-code/README_en.md

@@ -28,7 +28,7 @@ The additional details of the Aquila model will be presented in the official tec
 | AquilaCode-7B-TS   | Base model, "text-code" generation model, further pre-trained based on Aquila-7B, trained on Horizon Robotics chips | Same as above | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code) | [Download AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | Released        | Tianshu-BI-V100 |
 
 
-We will continue to release improved versions of Aquila model as open source. You can start by deleting the `model_pytorch.bi`n file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
+We will continue to release improved versions of Aquila model as open source. You can start by deleting the model checkpoint file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
 
  <br>If you have any question, please refer to the [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371) first. If you cannot solve them, please submit an [issue](https://github.com/FlagAI-Open/FlagAI/issues) directly.
 
@@ -145,7 +145,7 @@ For the above examples, you can modify the following parameters to achieve diffe
 | warm_up        | float | The ratio of the initial learning rate to the original learning rate. |
 | save_interval  | int   | The interval at which the model is saved, that is, how often the model is saved every few iterations of training. When the training time is long, the save interval can prevent all training results from being lost due to sudden interruptions or errors. |
 | log_interval   | int   | The interval at which logs are output, that is, how often log information is output every few iterations of training. |
-| lora           | int   | An integer value to enable LoRA optimization method during training. By default, it is set to 0 (no LoRA).|
+| lora           |bool   | Whether to enable LoRA optimization method during training. By default, it is set to False (no LoRA).|
 | enable_sft_dataset_dir | str | The directory of the SFT training dataset. |
 | enable_sft_dataset_file | str | The file name of the SFT training dataset. 
 

examples/Aquila/Aquila-pretrain/README.md

@@ -25,7 +25,7 @@
 | AquilaCode-7B-NV          | 基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在英伟达芯片完成训练  |   AquilaCode-7B 以小数据集、小参数量，实现高性能，是目前支持中英双语的、性能最好的开源代码模型，经过了高质量过滤、使用有合规开源许可的训练代码数据进行训练。<br><br> AquilaCode-7B 分别在英伟达和国产芯片上完成了代码模型的训练。  | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  |[下载AquilaCode-7B-NV](https://model.baai.ac.cn/model-detail/100102)  | 已发布  | Nvidia-A100 | 
 | AquilaCode-7B-TS           |基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在天数智芯芯片上完成训练  |    同上    | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  | [下载AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | 已发布  | Tianshu-BI-V100  | 
 
-悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的 `model_pytorch.bin`，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)** 。
+悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的 `checkpoints_in/aquila-7b`，再下载新权重，其他使用方式不变。详情见：**[变更日志](../changelog_zh.md)** 。
 
 <br>如有使用问题请先查看 [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371)，若不能解决，请直接提交 [issue](https://github.com/FlagAI-Open/FlagAI/issues) ~
 

examples/Aquila/Aquila-pretrain/README_en.md

@@ -27,7 +27,7 @@ The additional details of the Aquila model will be presented in the official tec
 | AquilaCode-7B-TS   | Base model, "text-code" generation model, further pre-trained based on Aquila-7B, trained on Horizon Robotics chips | Same as above | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code) | [Download AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | Released        | Tianshu-BI-V100 |
 
 
-We will continue to release improved versions of Aquila model as open source. You can start by deleting the `model_pytorch.bi`n file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
+We will continue to release improved versions of Aquila model as open source. You can start by deleting the model checkpoint file in the original directory and then download the new weights. Other usage methods remain unchanged. For more details, please refer to the **[Change Log](../changelog.md)**.
 
  <br>If you have any question, please refer to the [FAQ](https://github.com/FlagAI-Open/FlagAI/issues/371) first. If you cannot solve them, please submit an [issue](https://github.com/FlagAI-Open/FlagAI/issues) directly.
 
@@ -132,6 +132,9 @@ Note: The Aquila-7B base model may not perform as well for dialogue reasoning ta
     ```
     bash dist_trigger_docker.sh hostfile Aquila-chat-lora.yaml aquila-7b aquila_experiment
     ```
+    Note: When training Lora, it will generate an `adapter_config.json` and `adapter_model.bin` file, located in the output directory (at the same level as the log file). For inference, please run the `Aquila-chat/generate_chat_lora.py` file. The difference compared to regular inference is that the autoloader, when loading the model for inference, requires specifying the directory of the adapter files in the `adapter_dir` parameter.
+
+
 
 <details><summary>The correct output information is shown below:</summary>
 
@@ -171,7 +174,7 @@ For the above examples, you can modify the following parameters to achieve diffe
 | warm_up        | float | The ratio of the initial learning rate to the original learning rate. |
 | save_interval  | int   | The interval at which the model is saved, that is, how often the model is saved every few iterations of training. When the training time is long, the save interval can prevent all training results from being lost due to sudden interruptions or errors. |
 | log_interval   | int   | The interval at which logs are output, that is, how often log information is output every few iterations of training. |
-| lora           | int   | An integer value to enable LoRA optimization method during training. By default, it is set to 0 (no LoRA).|
+| lora           |bool   | Whether to enable LoRA optimization method during training. By default, it is set to False (no LoRA).|
 | enable_sft_dataset_dir | str | The directory of the SFT training dataset. |
 | enable_sft_dataset_file | str | The file name of the SFT training dataset. 
 

examples/Aquila/README.md

@@ -26,7 +26,7 @@
 | AquilaCode-7B-NV          | 基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在英伟达芯片完成训练  |   AquilaCode-7B 以小数据集、小参数量，实现高性能，是目前支持中英双语的、性能最好的开源代码模型，经过了高质量过滤、使用有合规开源许可的训练代码数据进行训练。<br><br> AquilaCode-7B 分别在英伟达和国产芯片上完成了代码模型的训练。  | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  |[下载AquilaCode-7B-NV](https://model.baai.ac.cn/model-detail/100102)  | 已发布  | Nvidia-A100 | 
 | AquilaCode-7B-TS           |基础模型，“文本-代码”生成模型，基于 Aquila-7B继续预训练，在天数智芯芯片上完成训练  |    同上    | [./examples/Aquila/Aquila-code](https://github.com/FlagAI-Open/FlagAI/tree/master/examples/Aquila/Aquila-code)  | [下载AquilaCode-7B-TS](https://model.baai.ac.cn/model-detail/100099)  | 已发布  | Tianshu-BI-V100  | 
 
-悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的 `model_pytorch.bin`，再下载新权重，其他使用方式不变。详情见：**[变更日志](./changelog_zh.md)** 。
+悟道·天鹰Aquila系列模型将持续开源更优版本，大家可以先删除原来目录下的`checkpoints_in/aquila-7b`，再下载新权重，其他使用方式不变。详情见：**[变更日志](./changelog_zh.md)** 。
 
 
 
@@ -186,7 +186,7 @@ python generate_bminf.py
 | warm_up           |float |   初始学习率与原始学习率的比例;     | 
 | save_interval         | int  |   模型保存的间隔，即每训练多少个iteration保存一次模型。当训练时间较长时，保存间隔可以避免因突然中断或出现错误导致训练成果全部丢失;   | 
 | log_interval           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
-| lora           |int  |    日志输出的间隔，即每训练多少个iteration输出一次日志信息    | 
+| lora           |bool  |    是否启用lora微调   | 
 | enable_sft_dataset_dir           |str  |    SFT训练数据集的目录息    | 
 | enable_sft_dataset_file           |str  |    SFT训练数据集的文件名    | 
 

examples/Aquila/README_en.md

@@ -132,6 +132,7 @@ Note: The Aquila-7B basic model may not perform as well for dialogue reasoning t
     ```
     bash dist_trigger_docker.sh hostfile Aquila-chat-lora.yaml aquila-7b aquila_experiment
     ```
+    Note: When training Lora, it will generate an `adapter_config.json` and `adapter_model.bin` file, located in the output directory (at the same level as the log file). For inference, please run the `Aquila-chat/generate_chat_lora.py` file. The difference compared to regular inference is that the autoloader, when loading the model for inference, requires specifying the directory of the adapter files in the `adapter_dir` parameter.
 
 <details><summary>The correct output information is shown below:</summary>
 
@@ -171,7 +172,7 @@ For the above examples, you can modify the following parameters to achieve diffe
 | warm_up        | float | The ratio of the initial learning rate to the original learning rate. |
 | save_interval  | int   | The interval at which the model is saved, that is, how often the model is saved every few iterations of training. When the training time is long, the save interval can prevent all training results from being lost due to sudden interruptions or errors. |
 | log_interval   | int   | The interval at which logs are output, that is, how often log information is output every few iterations of training. |
-| lora           | int   | An integer value to enable LoRA optimization method during training. By default, it is set to 0 (no LoRA).|
+| lora           |bool   | Whether to enable LoRA optimization method during training. By default, it is set to False (no LoRA).|
 | enable_sft_dataset_dir | str | The directory of the SFT training dataset. |
 | enable_sft_dataset_file | str | The file name of the SFT training dataset.