deep-text-recognition-benchmark issues

can not load pretrain model

1

I use nvidia 3090, cuda 11.3, py3.6 torch==1.10.0+cu113 torchvision==0.11.1+cu113 torchaudio==0.10.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html currently I still can not load pretrain model to continue training please help me.. log is here: Total_batch_size:...

gvo141

support training progress bar

Support of displaying a progress bar while training

bakrianoo

set new parameter to save the model to a specific path

Support of specifying the path where the model will be saved, rather than saving it to a relative path.

bakrianoo

cost is nan

what is the problem about cost is nan?

LinXin04

Use your pretrained model and finetune on my own data

7

Hello, How do I use your pretrained model and finetune on my own data. I downloaded your models and arranged my data as suggested. Should I use python train.py --saved_model...

janetwise

Pretrained model hyperparameters

4

Hi @ku21fan Just want to know which hyperparameters (no. of iterations, lr, batch size, optimizer) are used to train TPS-ResNet-BiLSTM-Attn-case-sensitive model? also what was the final training loss and validation...

ladanisavan

The link of lmdb dataset is invalid

https://www.dropbox.com/sh/i39abvnefllx2si/AAAbAYRvxzRp3cIE5HzqUw3ra?dl=0 this link report 404

czr9775

Unable to download the LMDB dataset in the example

3

I'm trying to download the dataset mentioned [here](https://github.com/clovaai/deep-text-recognition-benchmark#download-lmdb-dataset-for-traininig-and-evaluation-from-here) and I keep getting `The zip file is too large.`

ghost