You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When reducing the training set characters, an error is reported: File "/home/jmt/MXfont/mxfont-main/trainer/trainer_utils.py", line 101, in expert_assign r_in, c_in = linear_sum_assignment(prob_in)
ValueError: matrix contains invalid numeric entrie.
There should be a free data matrix, but increasing the batchsize will not cause an error, but I don't want to increase the batchsize. May I ask if there is any solution available
The text was updated successfully, but these errors were encountered:
My data is fine, it only uses 556 characters. May I ask what is causing this? I have been locating it for a long time and cannot solve it. Can you provide the following assistance
It seems that the training is unstable. "ValueError: matrix contains invalid numeric entrie." means that the input contains NaN, which can caused by various factors. I cannot know what the reason exactly is, but I guess using less characters than our setting may cause unstable training.
------------------ 原始邮件 ------------------
发件人: "Song ***@***.***>;
发送时间: 2023年11月21日(星期二) 中午11:44
收件人: ***@***.***>;
抄送: ***@***.***>; ***@***.***>;
主题: Re: [clovaai/mxfont] When reducing the training set characters, an error is reported: (Issue #32)
It seems that the training is unstable. "ValueError: matrix contains invalid numeric entrie." means that the input contains NaN, which can caused by various factors. I cannot know what the reason exactly is, but I guess using less characters than our setting may cause unstable training.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
When reducing the training set characters, an error is reported: File "/home/jmt/MXfont/mxfont-main/trainer/trainer_utils.py", line 101, in expert_assign r_in, c_in = linear_sum_assignment(prob_in)
ValueError: matrix contains invalid numeric entrie.
There should be a free data matrix, but increasing the batchsize will not cause an error, but I don't want to increase the batchsize. May I ask if there is any solution available
The text was updated successfully, but these errors were encountered: