Layout-Parser layout-model-training issues

Layout-Parser / layout-model-training

The scripts for training Detectron2-based Layout Models on popular layout analysis datasets

202 stars 55 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Unable to reproduce performance

#26 y2sman opened 5 months ago
0
Not able to fetch all tables and figures when converting pdf into images

#25 reema93jain opened 9 months ago
0
adjust parameters like the **iteration number** ?

#24 MachineLearning4Work opened 9 months ago
1
Predicted boxes or scores contain Inf/NaN. Training has diverged.

#23 MachineLearning4Work opened 9 months ago
1
after training bash train_prima.sh, model_final.pth is not getting generated into output folder.

#22 ShaileshSardaTTL opened 1 year ago
0
FloatingPointError: Predicted boxes or scores contain Inf/NaN. Training has diverged.

#21 ShaileshSardaTTL closed 1 year ago
2
Scripts for training Mask-RCNN and Fast-RCNN

#20 Mukilan-Krishnakumar closed 1 year ago
1
Config File Publaynet fastrcnn _pickle.UnpicklingError: invalid load key, '<'. error

#19 Niteshiya opened 1 year ago
0
Issue on WandB and ClearML not tracking the metrics of Fast RCNN while training

#18 ammarsaf opened 1 year ago
0
How to decide (or know) the epoch of training?

#17 ammarsaf closed 1 year ago
0
Is there any way to save the progress of model training at each iteration ?

#16 Ashish45351 opened 1 year ago
1
Huge time taken to finetune model on 10 samples

#15 karthikgali opened 1 year ago
0
No module named 'funcy'error is coming in cocosplit.py

#14 gopaljadhavv closed 1 year ago
0
TypeError: object of type 'int' has no len()

#13 gamingflexer opened 1 year ago
0
Using Detectron2 on CPU

#12 jan-schaeffer opened 2 years ago
3
Loss becoming infinite

#11 Mohdwajtech opened 2 years ago
1
Add fine-tuning in training

#10 lolipopshock opened 2 years ago
0
alter for multi-gpu training

#9 TITC closed 2 years ago
1
KeyError: "Dataset 'xxxx-train' is not registered!

#8 TITC closed 2 years ago
0
fix: ValueError: With n_samples=0, test_size=None and train_size=0.8, the resulting train set will be empty. Adjust any of the aforementioned parameters.

#7 XBeg9 closed 2 years ago
2
Need suggestion while label mapping

#6 SAIVENKATARAJU opened 2 years ago
2
custom model

#5 chris-ml92 opened 3 years ago
2
Potential issue with cocosplit

#4 bholmessyapse closed 3 years ago
2
Issue with the coco split utility failing when info and license fields aren't available.

#3 nasheedyasin closed 3 years ago
2
Fixing some issues with coco split utility

#2 nasheedyasin closed 3 years ago
2
support pre-training or fine-tuning schemes

#1 bertsky opened 4 years ago
18