YOLO
YOLO
> 是从0 开始啊,只是不冻结了 好的,谢谢。 另外想问一下,epoch0的时候loss到四位数这样正常吗?好像和最后差的有点大了 画出来的loss图有点奇怪
loss 从1000左右到20左右,现在的loss图是这样 
> ### 自我尝试 > 不管什么问题,请先尝试自行解决,“万般努力”之下仍然无法解决再来提问。此处请贴上你的努力过程。 import tensorflow.compat.v1 as tf tf.disable_v2_behavior() > > import keras.backend.tensorflow_backend as tb tb._SYMBOLIC_SCOPE.value = True > > 降低keras版本都不行 您好,请问您这个问题解决了吗?
> Okay!And thanks for your great code work, this is a very clear template!
Hi,did you delete your comment?I can only see it in the email notification,not in the githun issue 发自我的iPhone ------------------ Original ------------------ From: Simon Diener ***@***.***> Date: Tue,Jul 26,2022 3:35 AM...
Thank you for your reply~ After I looked at the code, I relearned it. I wonder if it can be understood as follows: MSRF-Net does not have a Canny operation...
> I mean that I think we can just use x13, X23, X33, x44 to get what we want without concatenating the edges of the original image;Maybe I was wrong...
Hi, what should I do with the .mat file? Why are the elements in the matrix I read negative?😭
Thank you for your reply Where can I find 'oasbud/raw_rf' folders and 'oasbud/b_mode'folders ? ------------------ 原始邮件 ------------------ 发件人: "tensorflow/datasets" ***@***.***>; 发送时间: 2022年7月14日(星期四) 上午6:58 ***@***.***>; ***@***.******@***.***>; 主题: Re: [tensorflow/datasets] [data request] OASBUD (#2371) Hi, what...
> 我发现最严重的错误应该在于trainer.py的训练函数和验证函数中的Sentence LTSM没有使用 > `prev_hidden_states = hidden_state`更新 > 我还对损失函数和预处理时的图像变换做了一些修改,现在它可以输出不同的句子了,但看上去效果依然不是很理想 > 我之后应该可以放出我的代码 您好,请问您放出代码了吗?想要参考学习一下