未验证 提交 8a5ba302 编写于 作者: M Meiyim 提交者: GitHub

update seq2seq readme (#485)

* update seq2seq readme

* update readme

* Update README.zh.md
上级 0ad3a7f0
...@@ -179,7 +179,7 @@ many other demo python scripts: ...@@ -179,7 +179,7 @@ many other demo python scripts:
1. [Semantic Similarity](./demo/finetune_classifier_dygraph.py) 1. [Semantic Similarity](./demo/finetune_classifier_dygraph.py)
1. [Name Entity Recognition(NER)](./demo/finetune_ner_dygraph.py) 1. [Name Entity Recognition(NER)](./demo/finetune_ner_dygraph.py)
1. [Machine Reading Comprehension](./demo/finetune_mrc_dygraph.py) 1. [Machine Reading Comprehension](./demo/finetune_mrc_dygraph.py)
1. [Text generation](./experimental/seq2seq/README.md) 1. [Text generation](./demo/seq2seq/README.md)
......
...@@ -80,7 +80,7 @@ print(pooled.numpy()) # convert results to numpy ...@@ -80,7 +80,7 @@ print(pooled.numpy()) # convert results to numpy
本项目依赖PaddlePaddle 1.7.0+, 请参考[这里](https://www.paddlepaddle.org.cn/install/quick)安装 PaddlePaddle。 本项目依赖PaddlePaddle 1.7.0+, 请参考[这里](https://www.paddlepaddle.org.cn/install/quick)安装 PaddlePaddle。
##### 2. 安装 ERNIE ##### 2. 安装 ERNIE 套件
```script ```script
...@@ -94,7 +94,10 @@ git clone https://github.com/PaddlePaddle/ERNIE.git --depth 1 ...@@ -94,7 +94,10 @@ git clone https://github.com/PaddlePaddle/ERNIE.git --depth 1
cd ERNIE cd ERNIE
pip install -r requirements.txt pip install -r requirements.txt
pip install -e . pip install -e .
```
`propeller`是辅助模型训练的高级框架,包含NLP常用的前、后处理流程。你可以通过将本repo根目录放入`PYTHONPATH`的方式导入`propeller`:
```shell
export PYTHONPATH=$PWD:$PYTHONPATH
``` ```
##### 3. 下载预训练模型(可选)<a name="section-pretrained-models"></a> ##### 3. 下载预训练模型(可选)<a name="section-pretrained-models"></a>
...@@ -179,7 +182,7 @@ python3 -m paddle.distributed.launch \ ...@@ -179,7 +182,7 @@ python3 -m paddle.distributed.launch \
1. [语义匹配](./demo/finetune_classifier_dygraph.py) 1. [语义匹配](./demo/finetune_classifier_dygraph.py)
1. [命名实体识别(NER)](./demo/finetune_ner_dygraph.py) 1. [命名实体识别(NER)](./demo/finetune_ner_dygraph.py)
1. [机器阅读理解](./demo/finetune_mrc_dygraph.py) (需要多卡环境运行;参见上面"分布式 finetune"一节) 1. [机器阅读理解](./demo/finetune_mrc_dygraph.py) (需要多卡环境运行;参见上面"分布式 finetune"一节)
1. [文本摘要生成](./experimental/seq2seq/README.md) 1. [文本摘要生成](./demo/seq2seq/README.md)
**推荐超参数设置:** **推荐超参数设置:**
......
...@@ -12,7 +12,7 @@ To starts finetuning ERNIE-GEN, run: ...@@ -12,7 +12,7 @@ To starts finetuning ERNIE-GEN, run:
```script ```script
python3 -m paddle.distributed.launch \ python3 -m paddle.distributed.launch \
--log_dir ./log \ --log_dir ./log \
./ernie_d/experimental/finetune_seq2seq_dygraph.py \ ./demo/seq2seq/finetune_seq2seq_dygraph.py \
--from_pretrained ernie-gen-base-en \ --from_pretrained ernie-gen-base-en \
--data_dir ./data/cnndm \ --data_dir ./data/cnndm \
--save_dir ./model_cnndm \ --save_dir ./model_cnndm \
...@@ -52,7 +52,7 @@ To run beam serach decode after you got a finetuned model. try: ...@@ -52,7 +52,7 @@ To run beam serach decode after you got a finetuned model. try:
```shell ```shell
cat one_column_source_text| python3 ernie_d/experimental/seq2seq/decode.py \ cat one_column_source_text| python3 demo/seq2seq/decode.py \
--from_pretrained ./ernie_gen_large \ --from_pretrained ./ernie_gen_large \
--save_dir ./model_cnndm \ --save_dir ./model_cnndm \
--bsz 8 --bsz 8
......
../demo/seq2seq
\ No newline at end of file
../../demo/seq2seq/README.md
\ No newline at end of file
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册