27号

Results 25 comments of 27号

3.6.5，host_name中文无法显示

![image](https://user-images.githubusercontent.com/2245040/87755616-e4784600-c839-11ea-9b0b-e7055a4b687f.png) 同样：中文乱码

3.6.5，host_name中文无法显示

版本号写错了。。3.5.6！

Support for JSON5

i want to use comments. thx. if i use comments and format it , then comments lost

连接本地xshell6，调用xftp，下载异常，其他操作正常

目前还不支持xshell中直接打开xftp软件，建议使用web页面的sftp进行文件的传输。

[QUESTION] fastjson1迁移至2的用法问题

version: 2.0.44 error: `The method toJSONString(Object, JSONWriter.Feature...) in the type JSONObject is not applicable for the arguments (List, SimplePropertyPreFilter)`

Error: /paddle/paddle/phi/kernels/gpu/bce_loss_kernel.cu:42 Assertion `(x >= static_cast<T>(0)) && (x <= one)` failed. Input is expected to be within the interval [0, 1], but received nan.

> 按照显卡数量和bs调整学习率可以解决 我贴的配置，就是调整过的。

Error: /paddle/paddle/phi/kernels/gpu/bce_loss_kernel.cu:42 Assertion `(x >= static_cast<T>(0)) && (x <= one)` failed. Input is expected to be within the interval [0, 1], but received nan.

> 你先试一下不用amp训练看看 @ghostxsl ，按你说的，去掉amp，还是报一样的错误。

Error: /paddle/paddle/phi/kernels/gpu/bce_loss_kernel.cu:42 Assertion `(x >= static_cast<T>(0)) && (x <= one)` failed. Input is expected to be within the interval [0, 1], but received nan.

不行，我更换到python3.9，也是报类似的错误。

Error: /paddle/paddle/phi/kernels/gpu/bce_loss_kernel.cu:42 Assertion `(x >= static_cast<T>(0)) && (x <= one)` failed. Input is expected to be within the interval [0, 1], but received nan.

> [#6723 (comment)](https://github.com/PaddlePaddle/PaddleDetection/issues/6723#issuecomment-1326083748) 你先试下这里的单测用例，看看是否在你的环境下也会出现类似的bug 我用这个代码测试了，没有得出同样的输出信息。多次执行，只返回如下信息： ``` Tensor(shape=[3], dtype=int64, place=Place(gpu:0), stop_gradient=True, [0, 1, 2]) ``` 看错误信息，他们好像不是一个地方产生的。

Error: /paddle/paddle/phi/kernels/gpu/bce_loss_kernel.cu:42 Assertion `(x >= static_cast<T>(0)) && (x <= one)` failed. Input is expected to be within the interval [0, 1], but received nan.

是不是和这个参数有关？ 默认的是： pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams 我使用的是： pretrain_weights: https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco.pdparams