search

zetane / viewer

ML models and internal tensors 3D visualizer
1.11k stars 122 forks source link

Environment Error #13

Closed wuyun-long closed 2 years ago

wuyun-long commented 2 years ago

When I load a keras file in .h5 format, an environment error is reported. Why?

konafah commented 2 years ago

Hello. Thanks for trying Zetane. Please try loading the h5 again, then closing the viewer, then sending us the latest zetane_error_log .

The logs can be found in:

You can also zip all the logs and email me directly at omar@zetane.com

wuyun-long commented 2 years ago

Thank you for your reply. In the attachment is the log file of the program, currently I encountered two kinds of problems 1. environment error, 2. value error. Thank you for your reply again! Yours sincerely, Wuyun-long

------------------ 原始邮件 ------------------ 发件人: "zetane/viewer" @.>; 发送时间: 2022年2月18日(星期五) 晚上10:27 @.>; @.**@.>; 主题: Re: [zetane/viewer] Environment Error (Issue #13)

Hello. Thanks for trying Zetane. Please try loading the h5 again, then closing the viewer, then sending us the latest zetane_error_log .

The logs can be found in:

macOS: ~/Library/Application\ Support/Zetane/logs

Ubuntu: ~/.config/Zetane/logs

Windows: %APPDATA%/Zetane/logs

You can also zip all the logs and email me directly at @.***

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.Message ID: @.***>

konafah commented 2 years ago

Thank you for the logs, here's what we saw:

  1. For the first error, it looks like pypi (files.pythonhosted.org) was not reachable from your network, so the keras environment setup was not able to complete. However, in later attempts (02-21-2022) it looks like the connectivity issue was resolved and the environment completed its setup correctly.

  2. The second error ValueError: No model config found in the file can happen when the h5 file only has weights. Try re-saving the whole model (configuration + weights) as suggested here: