-
【demo】semantic_role_labeling中的脚本,缺少可执行权限。
drwxr-xr-x 7 baidu staff 238B 12 5 13:29 data
-rw-r--r-- 1 baidu staff 2.1K 12 5 13:29 dataprovider.py
-rw-r--r-- 1 baidu staff 4.2K 12 5…
-
Hi, I have tried to train TE model with experiment_config/decomposable_attention.json, and I adjust the epoch number to 3000. but I have only got the best validation accuracy of 72.8%,with training ac…
-
The example of using a
[disclosure for answers in a FAQ](http://w3c.github.io/aria-practices/examples/disclosure/disclosure-faq.html)
is ready for review.
#### Requested reviews
- [x] Review…
-
It hangs when the last pass is trained, for the following settings:
1. use_gpu = false, trainer_count=4
2. use_gpu = true, trainer_count=4
3. use_gpu = true, trainer_count=1
It can exit normally onl…
-
1. demo/semantic_role_labeling中的可执行脚本需要一可执行方式更新到github中,当前demo中并没有执行权限,建议以执行权限更新。
![image](https://cloud.githubusercontent.com/assets/6229206/20453866/2a50186e-ae6d-11e6-99de-7b31a5c7c6db.png)
![ima…
-
这里只训练2个pass,训练时test accuracy 是0.552631. 但是model reload进来以后test accuracy是0.552867. 具体log信息如下:
==training log==
......I1202 14:09:10.074734 20597 TrainerInternal.cpp:182] Pass=1 Batch=36 samples=52…
-
Hi,
I tried to run the srl demo but I could not download the embeddings using https://github.com/PaddlePaddle/Paddle/blob/develop/demo/semantic_role_labeling/data/get_data.sh.
I got error 404 from g…
enfry updated
7 years ago
-
Instead of using classes, mini.css could make any site that uses it more accessible CSS by default if you use [ARIA roles](https://www.w3.org/TR/html-aria/) to identify tabs, buttons, alerts, menus an…
-
1. 在使用demo中test.sh脚本时,代码中,function get_best_pass()方法的目的是找到cost最小的pass-xxx模型。但是在使用sort排序是没有加-n参数,所得到的结果并不一定是最小的。
例如,在semantic_role_labeling中,训练500个模型,按照test.sh中脚本执行方法是,最小的是cost=10.0364的模型 00376,但实际上有存…
-
我希望的训练预料来自不同文件,通过ID关联,或者拼接成一条文本也行。
每份预料对应不同的dict.txt
组成2个以上的slot。
例如:
0;0 1 4 1023;1 4 7 3;2 2 6 511;
1;2 1 5 88 511;
请问有类似的dataprovider可以作为参考么?