Issue Summary:
I'm having difficulty installing and using this package to process a single JPG file. As a non-developer, I've struggled to identify the problem. I've tried multiple approaches across different environments but have not succeeded. I tried to install this entire day.. Below are the details of the issues I've encountered:
ImportError: cannot import name 'EffOCR' from 'efficient_ocr' (unknown location)
The effocr.py file is located under the src folder, which seems to be causing issues with initializing the package. The two-line code example provided did not work for me.
Windows:
I tried installing the package on two different Windows machines. In both cases, after running the import statement, the kernel crashed:
!git clone https://github.com/dell-research-harvard/efficient_ocr.git
%cd efficient_ocr
from efficient_ocr import EffOCR
Mac (M1 chip):
On my Mac with an M1 chip, I was forced to download the CPU version of FAISS since it doesn't support the M1 architecture. However, when trying to process an image, I encountered the following error:
TypeError: 'tuple' object does not support item assignment
Package Confusion:
After multiple failed attempts, I found the current GitHub page (https://github.com/dell-research-harvard/effocr). However, I noticed that it does not support the EffOCR import as mentioned in the other documentations.
Request:
Could you please clarify which package source I should be using? Additionally, are there specific environment requirements I should be aware of to get the package running successfully?
Issue Summary:
I'm having difficulty installing and using this package to process a single JPG file. As a non-developer, I've struggled to identify the problem. I've tried multiple approaches across different environments but have not succeeded. I tried to install this entire day.. Below are the details of the issues I've encountered:
Google Colab:
I attempted to use the package on Google Colab(https://colab.research.google.com/drive/1cbd4Hq5V7lI6vvhQ8u8nm_jKU_LBdyVw?usp=sharing), but encountered the following error:
The
effocr.py
file is located under thesrc
folder, which seems to be causing issues with initializing the package. The two-line code example provided did not work for me.Windows:
I tried installing the package on two different Windows machines. In both cases, after running the import statement, the kernel crashed:
Mac (M1 chip):
On my Mac with an M1 chip, I was forced to download the CPU version of FAISS since it doesn't support the M1 architecture. However, when trying to process an image, I encountered the following error:
Package Confusion:
After multiple failed attempts, I found the current GitHub page (https://github.com/dell-research-harvard/effocr). However, I noticed that it does not support the
EffOCR
import as mentioned in the other documentations.Request:
Could you please clarify which package source I should be using? Additionally, are there specific environment requirements I should be aware of to get the package running successfully?
Thank you for your assistance.