Ramsri Goutham Golla comments

Results 7 comments of


                                            Ramsri Goutham Golla

"WARNING:transformers.tokenization_utils_base:Token indices ..." comes with main.BoolQGen() and main.AnswerPredictor()

Hi @alexcruz0202 , Split the text into smaller chunks. Keep each chunk around 200-300 words at the maximum.

Improve segmentation step to get single label and single marker for each object

Thanks @SkalskiP Is there support/potential to include support for https://github.com/UX-Decoder/Segment-Everything-Everywhere-All-At-Once in this library? Basically, I was trying to get better segmentation masks instead of traditional SAM and merging which is...

Improve segmentation step to get single label and single marker for each object

Understood thanks for the quick response!

Unable to run train.py

Google colab should work fine for training this model. My advice would be to try that if local GPU is causing issues.

Unable to run train.py

Hi @mohammad-sameer-khan , Mostly there might be memory-related issues of GPU. Try to reduce batch size etc. Check few t5-finetuning colab notebooks like this one - https://colab.research.google.com/drive/176NSaYjc2eeI-78oLH_F9-YV3po3qQQO?usp=sharing#scrollTo=HVxGfmEMCKs_

Running it in Colab

Thanks @abhi1nandy2 !

Not bugs but customisations

Please take a look at the following videos to precisely control text position: https://youtu.be/zvORP8LCLC4 https://youtu.be/LWrRJx2wdWc https://youtu.be/Ex1kuxe6jRg You can move text to any position. Regards, Ramsri On Mon, 21 Aug 2023...