EasyOCR icon indicating copy to clipboard operation
EasyOCR copied to clipboard

Error in loading tamil model

Open StaronDsilva opened this issue 1 year ago • 5 comments

Tried to load tamil model but faced this error:

raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(

RuntimeError: Error(s) in loading state_dict for Model: size mismatch for Prediction.weight: copying a param with shape torch.Size([143, 512]) from checkpoint, the shape in current model is torch.Size([127, 512]). size mismatch for Prediction.bias: copying a param with shape torch.Size([143]) from checkpoint, the shape in current model is torch.Size([127]).

Please try to provide the solution .Thanks

StaronDsilva avatar Sep 06 '23 04:09 StaronDsilva

Using version 1.6.2 can fixed this issues

passerbya avatar Sep 27 '23 08:09 passerbya

I have a similar issue even on 1.6.2. Also tried 1.7.1 - same error.

RuntimeError: Error(s) in loading state_dict for DataParallel:
        size mismatch for module.Prediction.weight: copying a param with shape torch.Size([103, 256]) from checkpoint, the shape in current model is torch.Size([97, 256]).
        size mismatch for module.Prediction.bias: copying a param with shape torch.Size([103]) from checkpoint, the shape in current model is torch.Size([97]).

MikeAtom avatar Oct 23 '23 01:10 MikeAtom

Try downgrading easyocr version, it helped me

On Mon, 23 Oct 2023, 6:49 am MikeAtom, @.***> wrote:

I have a similar issue even on 1.6.2. Also tried 1.7.1 - same error.

RuntimeError: Error(s) in loading state_dict for DataParallel: size mismatch for module.Prediction.weight: copying a param with shape torch.Size([103, 256]) from checkpoint, the shape in current model is torch.Size([97, 256]). size mismatch for module.Prediction.bias: copying a param with shape torch.Size([103]) from checkpoint, the shape in current model is torch.Size([97]).

— Reply to this email directly, view it on GitHub https://github.com/JaidedAI/EasyOCR/issues/1135#issuecomment-1774278379, or unsubscribe https://github.com/notifications/unsubscribe-auth/BBHVMXYB6HI7NISHQPYORD3YAXAYXAVCNFSM6AAAAAA4MW2B6OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONZUGI3TQMZXHE . You are receiving this because you authored the thread.Message ID: @.***>

StaronDsilva avatar Oct 23 '23 02:10 StaronDsilva

Hello MikeAtom

I also faced same issue change the yaml file as per opt.txt file and also in ~/dist-packages/easyor/character/your_language_char.txt need to be updated as per opt.txt character list

This error will be resolved

vamshi-1611 avatar Jan 23 '24 12:01 vamshi-1611

Please check this repo if you are interested in tamil ocr. It's capable of recognizing both tamil and English. https://github.com/gnana70/tamil_ocr

You can simply install using pip install ocr_tamil

from ocr_tamil.ocr import OCR
image_path = r"test_images\0.jpg" # insert your own path here
ocr = OCR(detect=True)
texts = ocr.predict(image_path)
with open("output.txt","w",encoding="utf-8") as f:
    f.write(texts)

>>>> கொடைக்கானல் Kodaikanal 

gnana70 avatar Jan 30 '24 21:01 gnana70