reciept-model-2500
This model is a fine-tuned version of microsoft/layoutlmv3-base on the format_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0043
- Precision: 0.9673
- Recall: 0.9625
- F1: 0.9649
- Accuracy: 0.9993
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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2500
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 0.62 | 100 | 0.0150 | 0.8575 | 0.8725 | 0.8649 | 0.9972 |
No log | 1.25 | 200 | 0.0075 | 0.8756 | 0.9325 | 0.9031 | 0.9979 |
No log | 1.88 | 300 | 0.0154 | 0.8744 | 0.8875 | 0.8809 | 0.9973 |
No log | 2.5 | 400 | 0.0118 | 0.8881 | 0.9525 | 0.9192 | 0.9982 |
0.0029 | 3.12 | 500 | 0.0091 | 0.9158 | 0.925 | 0.9204 | 0.9983 |
0.0029 | 3.75 | 600 | 0.0167 | 0.8720 | 0.9025 | 0.8870 | 0.9975 |
0.0029 | 4.38 | 700 | 0.0092 | 0.9183 | 0.9275 | 0.9229 | 0.9983 |
0.0029 | 5.0 | 800 | 0.0113 | 0.8843 | 0.9175 | 0.9006 | 0.9979 |
0.0029 | 5.62 | 900 | 0.0106 | 0.9349 | 0.8975 | 0.9158 | 0.9982 |
0.0017 | 6.25 | 1000 | 0.0043 | 0.9673 | 0.9625 | 0.9649 | 0.9993 |
0.0017 | 6.88 | 1100 | 0.0044 | 0.9602 | 0.965 | 0.9626 | 0.9993 |
0.0017 | 7.5 | 1200 | 0.0118 | 0.9246 | 0.92 | 0.9223 | 0.9982 |
0.0017 | 8.12 | 1300 | 0.0067 | 0.9406 | 0.95 | 0.9453 | 0.9988 |
0.0017 | 8.75 | 1400 | 0.0083 | 0.9409 | 0.955 | 0.9479 | 0.9989 |
0.001 | 9.38 | 1500 | 0.0060 | 0.9495 | 0.94 | 0.9447 | 0.9988 |
0.001 | 10.0 | 1600 | 0.0078 | 0.9369 | 0.9275 | 0.9322 | 0.9985 |
0.001 | 10.62 | 1700 | 0.0093 | 0.9248 | 0.9525 | 0.9384 | 0.9986 |
0.001 | 11.25 | 1800 | 0.0097 | 0.9062 | 0.9425 | 0.9240 | 0.9983 |
0.001 | 11.88 | 1900 | 0.0100 | 0.9098 | 0.9325 | 0.9210 | 0.9982 |
0.0006 | 12.5 | 2000 | 0.0111 | 0.9113 | 0.925 | 0.9181 | 0.9981 |
0.0006 | 13.12 | 2100 | 0.0107 | 0.9275 | 0.9275 | 0.9275 | 0.9983 |
0.0006 | 13.75 | 2200 | 0.0105 | 0.9279 | 0.9325 | 0.9302 | 0.9984 |
0.0006 | 14.38 | 2300 | 0.0109 | 0.9325 | 0.9325 | 0.9325 | 0.9985 |
0.0006 | 15.0 | 2400 | 0.0109 | 0.9325 | 0.9325 | 0.9325 | 0.9985 |
0.0003 | 15.62 | 2500 | 0.0109 | 0.9325 | 0.9325 | 0.9325 | 0.9985 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
- Downloads last month
- 2
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.
Model tree for akseljoonas/reciept-model-2500
Base model
microsoft/layoutlmv3-baseEvaluation results
- Precision on format_datasettest set self-reported0.967
- Recall on format_datasettest set self-reported0.963
- F1 on format_datasettest set self-reported0.965
- Accuracy on format_datasettest set self-reported0.999