layoutlmv3-finetuned-cord_300

This model is a fine-tuned version of microsoft/layoutlmv3-base on the cord-layoutlmv3 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3434
  • Precision: 0.9325
  • Recall: 0.9416
  • F1: 0.9371
  • Accuracy: 0.9363

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-05
  • train_batch_size: 5
  • eval_batch_size: 5
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 4000

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 4.17 250 1.0379 0.7204 0.7829 0.7504 0.7941
1.4162 8.33 500 0.5642 0.8462 0.8772 0.8614 0.8820
1.4162 12.5 750 0.3836 0.9055 0.9184 0.9119 0.9206
0.3211 16.67 1000 0.3209 0.9139 0.9296 0.9217 0.9334
0.3211 20.83 1250 0.2962 0.9275 0.9386 0.9330 0.9435
0.1191 25.0 1500 0.2979 0.9254 0.9379 0.9316 0.9402
0.1191 29.17 1750 0.3079 0.9282 0.9386 0.9334 0.9355
0.059 33.33 2000 0.3039 0.9232 0.9364 0.9298 0.9325
0.059 37.5 2250 0.3254 0.9248 0.9386 0.9316 0.9355
0.0342 41.67 2500 0.3404 0.9246 0.9364 0.9305 0.9334
0.0342 45.83 2750 0.3386 0.9354 0.9431 0.9392 0.9355
0.0226 50.0 3000 0.3274 0.9354 0.9431 0.9392 0.9359
0.0226 54.17 3250 0.3282 0.9341 0.9446 0.9393 0.9393
0.017 58.33 3500 0.3475 0.9319 0.9424 0.9371 0.9363
0.017 62.5 3750 0.3367 0.9340 0.9431 0.9385 0.9372
0.0145 66.67 4000 0.3434 0.9325 0.9416 0.9371 0.9363

Framework versions

  • Transformers 4.21.2
  • Pytorch 1.12.1+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
9
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Evaluation results