[Docs]: Translate 3_pretrain.md to Chinese (#552)
* [Docs]: Translate 3_pretrain.md to Chinese * [Docs]: adjust the details * [Fix]: fix lintpull/582/head
RenQin
Yixiao Fang
parent
5e56d2a1e1
commit
536ac8f448
|
|
@ -1,97 +1,96 @@
|
|||
# Tutorial 3: Pretrain with existing models
|
||||
# 教程 3: 使用现有模型进行预训练
|
||||
|
||||
- [Tutorial 3: Pretrain with existing models](#tutorial-3-pretrain-with-existing-models)
|
||||
- [Start to Train](#start-to-train)
|
||||
- [Train with a single GPU](#train-with-a-single-gpu)
|
||||
- [Train with CPU](#train-with-cpu)
|
||||
- [Train with multiple GPUs](#train-with-multiple-gpus)
|
||||
- [Train with multiple machines](#train-with-multiple-machines)
|
||||
- [Launch multiple jobs on a single machine](#launch-multiple-jobs-on-a-single-machine)
|
||||
- [教程 3: 使用现有模型进行预训练](#教程-3-使用现有模型进行预训练)
|
||||
- [使用单卡训练](#使用单卡训练)
|
||||
- [使用 CPU 训练](#使用-cpu-训练)
|
||||
- [使用多卡训练](#使用多卡训练)
|
||||
- [使用用多台机器训练](#使用用多台机器训练)
|
||||
- [在一台机器上启动多个任务](#在一台机器上启动多个任务)
|
||||
|
||||
This page provides basic usage about how to run algorithms and how to use some tools in MMSelfSup. For installation instructions and date preparation, please refer to [install.md](install.md) and [prepare_data.md](prepare_data.md).
|
||||
本文档提供有关如何运行算法以及如何使用 MMSelfSup 中的一些工具的基本用法。有关安装说明和数据准备，请参阅 [install.md](install.md) 和 [prepare_data.md](prepare_data.md)。
|
||||
|
||||
## Start to Train
|
||||
## 开始训练
|
||||
|
||||
**Note**: The default learning rate in config files is for specific number of GPUs, which is indicated in the config names. If using different number GPUs, the total batch size will change in proportion, you have to scale the learning rate following `new_lr = old_lr * new_ngpus / old_ngpus`.
|
||||
**注意**：配置文件中的默认学习率是针对特定数量的 GPU（GPU数量已在配置文件名称中注明）。如果使用不同数量的 GPUs，总的 batch size 将按比例变化，您必须按照 `new_lr = old_lr * new_ngpus / old_ngpus` 缩放学习率。
|
||||
|
||||
### Train with a single GPU
|
||||
### 使用单卡训练
|
||||
|
||||
```shell
|
||||
python tools/train.py ${CONFIG_FILE} [optional arguments]
|
||||
```
|
||||
|
||||
A simple example to start training:
|
||||
一个简单的例子来开启训练：
|
||||
|
||||
```shell
|
||||
python tools/train.py configs/selfsup/mae/mae_vit-base-p16_8xb512-coslr-400e_in1k.py
|
||||
```
|
||||
|
||||
### Train with CPU
|
||||
### 使用 CPU 训练
|
||||
|
||||
```shell
|
||||
export CUDA_VISIBLE_DEVICES=-1
|
||||
python tools/train.py ${CONFIG_FILE} [optional arguments]
|
||||
```
|
||||
|
||||
**Note**: We do not recommend users to use CPU for training because it is too slow. We support this feature to allow users to debug on machines without GPU for convenience.
|
||||
**注意**。我们不建议用户使用CPU进行训练，因为它太慢了。我们支持这个功能，是为了方便用户在没有GPU的机器上进行调试。
|
||||
|
||||
### Train with multiple GPUs
|
||||
### 使用多卡训练
|
||||
|
||||
```shell
|
||||
sh tools/dist_train.sh ${CONFIG_FILE} ${GPUS} [optional arguments]
|
||||
```
|
||||
|
||||
Optional arguments:
|
||||
可选参数：
|
||||
|
||||
- `--work-dir`: Inidicate your custom work directory to save checkpoints and logs.
|
||||
- `--resume`: Automatically find the latest checkpoint in your work directory. Or set `--resume ${CHECKPOINT_PATH}` to load the specific checkpoint file.
|
||||
- `--amp`: Enable automatic-mixed-precision training.
|
||||
- `--cfg-options`: Setting `--cfg-options` will modify the original configs. For example, setting `--cfg-options randomness.seed=0` will set seed for random number.
|
||||
- `--work-dir`：指示您的自定义工作目录以保存 checkpoints 和日志。
|
||||
- `--resume`：自动在你的工作目录中查找最新的 checkpoints。或者设置 `--resume ${CHECKPOINT_PATH}` 来加载特定的 checkpoints 文件。
|
||||
- `--amp`：启用自动混合精度训练。
|
||||
- `--cfg-options`：设置 `--cfg-options` 将修改原始配置。例如，设置 `--cfg-options randomness.seed=0` 将为随机数设置种子。
|
||||
|
||||
An example to start training with 8 GPUs:
|
||||
使用 8 个 GPUs 开始训练的示例：
|
||||
|
||||
```shell
|
||||
sh tools/dist_train.sh configs/selfsup/mae/mae_vit-base-p16_8xb512-coslr-400e_in1k.py 8
|
||||
```
|
||||
|
||||
Alternatively, if you run MMSelfSup on a cluster managed with **[slurm](https://slurm.schedmd.com/)**:
|
||||
或者，如果您在使用 **[slurm](https://slurm.schedmd.com/)** 管理的集群上运行 MMSelfSup：
|
||||
|
||||
```shell
|
||||
GPUS_PER_NODE=${GPUS_PER_NODE} GPUS=${GPUS} SRUN_ARGS=${SRUN_ARGS} sh tools/slurm_train.sh ${PARTITION} ${JOB_NAME} ${CONFIG_FILE} [optional arguments]
|
||||
```
|
||||
|
||||
An example to start training with 8 GPUs:
|
||||
使用 8 个 GPUs 开始训练的示例：
|
||||
|
||||
```shell
|
||||
# The default setting: GPUS_PER_NODE=8 GPUS=8
|
||||
# 默认设置: GPUS_PER_NODE=8 GPUS=8
|
||||
sh tools/slurm_train.sh Dummy Test_job configs/selfsup/mae/mae_vit-base-p16_8xb512-coslr-400e_in1k.py
|
||||
```
|
||||
|
||||
### Train with multiple machines
|
||||
### 使用用多台机器训练
|
||||
|
||||
If you launch with multiple machines simply connected with ethernet, you can simply run following commands:
|
||||
如果您想使用由以太网连接起来的多台机器， 您可以使用以下命令:
|
||||
|
||||
On the first machine:
|
||||
在第一台机器上:
|
||||
|
||||
```shell
|
||||
NNODES=2 NODE_RANK=0 PORT=${MASTER_PORT} MASTER_ADDR=${MASTER_ADDR} sh tools/dist_train.sh ${CONFIG} ${GPUS}
|
||||
```
|
||||
|
||||
On the second machine:
|
||||
在第二台机器上:
|
||||
|
||||
```shell
|
||||
NNODES=2 NODE_RANK=1 PORT=${MASTER_PORT} MASTER_ADDR=${MASTER_ADDR} sh tools/dist_train.sh ${CONFIG} ${GPUS}
|
||||
```
|
||||
|
||||
Usually it is slow if you do not have high speed networking like InfiniBand.
|
||||
但是，如果您不使用高速网路连接这几台机器的话，训练将会非常慢。
|
||||
|
||||
If you launch with **slurm**, the command is the same as that on single machine described above, but you need refer to [slurm_train.sh](https://github.com/open-mmlab/mmselfsup/blob/master/tools/slurm_train.sh) to set appropriate parameters and environment variables.
|
||||
如果您使用的是 **slurm** 来管理多台机器，您可以使用同在单台机器上一样的命令来启动任务，但是您必须得设置合适的环境变量和参数，具体可以参考 [slurm_train.sh](https://github.com/open-mmlab/mmselfsup/blob/master/tools/slurm_train.sh)。
|
||||
|
||||
### Launch multiple jobs on a single machine
|
||||
### 在一台机器上启动多个任务
|
||||
|
||||
If you launch multiple jobs on a single machine, e.g., 2 jobs of 4-GPU training on a machine with 8 GPUs, you need to specify different ports (29500 by default) for each job to avoid communication conflict.
|
||||
如果你在一台机器上启动多个任务，例如，在一台有 8 个 GPU 的机器上启动 2 个分别使用４块 GPU 的训练任务，你需要为每个任务指定不同的端口（默认为29500）以避免通信冲突。
|
||||
|
||||
If you use `dist_train.sh` to launch training jobs:
|
||||
如果你使用 `dist_train.sh` 来启动训练任务。
|
||||
|
||||
```shell
|
||||
CUDA_VISIBLE_DEVICES=0,1,2,3 PORT=29500 sh tools/dist_train.sh ${CONFIG_FILE} 4 --work-dir tmp_work_dir_1
|
||||
|
|
@ -99,23 +98,23 @@ CUDA_VISIBLE_DEVICES=0,1,2,3 PORT=29500 sh tools/dist_train.sh ${CONFIG_FILE} 4
|
|||
CUDA_VISIBLE_DEVICES=4,5,6,7 PORT=29501 sh tools/dist_train.sh ${CONFIG_FILE} 4 --work-dir tmp_work_dir_2
|
||||
```
|
||||
|
||||
If you use launch training jobs with slurm, you have two options to set different communication ports:
|
||||
如果你用 slurm 启动训练任务，你有两种方法来设置不同的通信端口。
|
||||
|
||||
Option 1:
|
||||
**方法 1 :**
|
||||
|
||||
In `config1.py`:
|
||||
在 `config1.py` 中:
|
||||
|
||||
```python
|
||||
env_cfg = dict(dist_cfg=dict(backend='nccl', port=29500))
|
||||
```
|
||||
|
||||
In `config2.py`:
|
||||
在 `config2.py` 中:
|
||||
|
||||
```python
|
||||
env_cfg = dict(dist_cfg=dict(backend='nccl', port=29501))
|
||||
```
|
||||
|
||||
Then you can launch two jobs with config1.py and config2.py.
|
||||
然后你可以用 config1.py 和 config2.py 启动两个任务。
|
||||
|
||||
```shell
|
||||
CUDA_VISIBLE_DEVICES=0,1,2,3 GPUS=4 sh tools/slurm_train.sh ${PARTITION} ${JOB_NAME} config1.py [optional arguments]
|
||||
|
|
@ -123,9 +122,9 @@ CUDA_VISIBLE_DEVICES=0,1,2,3 GPUS=4 sh tools/slurm_train.sh ${PARTITION} ${JOB_N
|
|||
CUDA_VISIBLE_DEVICES=4,5,6,7 GPUS=4 sh tools/slurm_train.sh ${PARTITION} ${JOB_NAME} config2.py [optional arguments]
|
||||
```
|
||||
|
||||
Option 2:
|
||||
**方法 2 :**
|
||||
|
||||
You can set different communication ports without the need to modify the configuration file, but have to set the `--cfg-options` to overwrite the default port in configuration file.
|
||||
你可以设置不同的通信端口，而不需要修改配置文件，但必须设置 `--cfg-options` 来覆盖配置文件中的默认端口。
|
||||
|
||||
```shell
|
||||
CUDA_VISIBLE_DEVICES=0,1,2,3 GPUS=4 sh tools/slurm_train.sh ${PARTITION} ${JOB_NAME} config1.py --work-dir tmp_work_dir_1 --cfg-options env_cfg.dist_cfg.port=29500
|
||||
|
|
|
|||
Loading…
Reference in New Issue