pkg version conflicts

[zhxiaokang]

The required versions of python==3.6.13 and pandas==1.3.0 are conflicted since pandas==1.3.0 requires python>=3.7, as reported by Conda.

I also tried to install pandas==1.3.0 by pip, but the latest available version is 1.1.5 (on date 19/6/2023)

[moloch12]

Hi, I also encountered this problem, and even after I installed the appropriate version, the model file still wouldn't load. Any suggestions?

Traceback (most recent call last): File "/public/home/moloch/be/new_analysis/ad_f3/deepbe/DeepBE/YE1-BE4max/DeepBE_YE1.py", line 9, in final_model = tf.keras.models.load_model('DeepBE_YE1_model',compile=False) File "/public/home/moloch/miniconda3/envs/py1/lib/python3.10/site-packages/keras/utils/traceback_utils.py", line 70, in error_handler raise e.with_traceback(filtered_tb) from None File "/public/home/moloch/miniconda3/envs/py1/lib/python3.10/site-packages/tensorflow/python/training/py_checkpoint_reader.py", line 66, in get_tensor return CheckpointReader.CheckpointReader_GetTensor( IndexError: Read less bytes than requested

[NahyeKim]

@sungchulc Could you please help with this?

[zhxiaokang]

Hi @moloch12 , after struggling with the environment and version issue for several days, I suddenly succeeded for once with even "inappropriate" verions of CUDA (12.0) and tensorflow (2.12.0). But it didn't work the next day.

It seems that the real reason behind your error IndexError: Read less bytes than requested (I also had the same and kept questioning about my environment setting) is because the model file cannot be fetched from LFS (the one in the repo *model/variables/variables.data-00000-of-00001 is just a pointer to a large file in the LFS), since the repository's data quota is used up, as I reported in this issue.

So as long as Issue 2 is resolved, your problem should be solved.

[sungchulc]

Hi guys I'll take a look at this issue and the issue 2 in a few days and get back to you guys. We'll recheck if LFS payment is up to date and if it still doesn't work, we'll try to provide an alternative way to download our models in a few days. Thanks

[moloch12]

@sungchulc Could you please help with this?

Thanks for developing this brilliant tool! I have tried your web version which is easy to use. However, I met this error which returns nothing

A PHP Error was encountered
Severity: Warning

Message: file_get_contents(data/77c6c908a69e5472d15e569964e206b2a1a2240f/DeepBE_Results.txt): failed to open stream: No such file or directory

Filename: views/result1.php

Line Number: 8

Backtrace:

File: /home/snusry/DeepCpf1/DeepBE/application/views/result1.php
Line: 8
Function: file_get_contents

File: /home/snusry/DeepCpf1/DeepBE/application/controllers/Job.php
Line: 210
Function: view

File: /home/snusry/DeepCpf1/DeepBE/index.php
Line: 315
Function: require_once

I would appreciate it if you and your team could fix the error. Best.

[sungchulc]

@moloch12 We believe that there has been some temporary unknown error at that time,, maybe possibly due to high traffic.. With the same sequence and options that you chose, we confirmed that the webtool was working (URL: http://deepcrispr.info/DeepBE/page/user/bf78f0316937a2d813858dd738513f6d009902ed). Please run it again and let us know if it works.

Also, please open new issues for errors that are not related to this current issue. Thanks

[moloch12]

@moloch12 We believe that there has been some temporary unknown error at that time,, maybe possibly due to high traffic.. With the same sequence and options that you chose, we confirmed that the webtool was working (URL: http://deepcrispr.info/DeepBE/page/user/bf78f0316937a2d813858dd738513f6d009902ed). Please run it again and let us know if it works.

Also, please open new issues for errors that are not related to this current issue. Thanks

Thank you. On my second try, it worked, and it looks like it was caused by the Internet. This tool is very robust and a simple refresh will work like a charm. Best.

[sungchulc]

I've just published a new release for the model files (codes are the same) since they are downloaded as git lfs pointers rather than actual files. Please download files model_v1a to j from the release note and merge/untar them for use.

Since the package versions used for development is quite outdated, I also checked if the models are runnable with the newer package versions. The models were runnable with the current google colab package versions as below.

TF = 2.12.0 Python = 3.10.12 Numpy = 1.22.4 Pandas = 1.5.3 cuda = 11.8 cudnn = 8.9

If there's still a problem with the model files or package versions, please let me know. Thanks