SPECTER-finetuned-DAGPap22
This model is a fine-tuned version of allenai/specter on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0023
- Accuracy: 0.9993
- F1: 0.9995
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: 6e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 0.3422 | 1.0 | 669 | 0.4135 | 0.8914 | 0.9140 |
| 0.1074 | 2.0 | 1338 | 0.1216 | 0.9746 | 0.9811 |
| 0.0329 | 3.0 | 2007 | 0.0064 | 0.9989 | 0.9992 |
| 0.0097 | 4.0 | 2676 | 0.0132 | 0.9972 | 0.9980 |
| 0.0123 | 5.0 | 3345 | 0.0231 | 0.9961 | 0.9971 |
| 0.0114 | 6.0 | 4014 | 0.0080 | 0.9985 | 0.9989 |
| 0.0029 | 7.0 | 4683 | 0.2207 | 0.9727 | 0.9797 |
| 0.0075 | 8.0 | 5352 | 0.0145 | 0.9974 | 0.9981 |
| 0.0098 | 9.0 | 6021 | 0.0047 | 0.9994 | 0.9996 |
| 0.0025 | 10.0 | 6690 | 0.0000 | 1.0 | 1.0 |
| 0.0044 | 11.0 | 7359 | 0.0035 | 0.9993 | 0.9995 |
| 0.0 | 12.0 | 8028 | 0.0027 | 0.9996 | 0.9997 |
| 0.0027 | 13.0 | 8697 | 0.0036 | 0.9993 | 0.9995 |
| 0.0055 | 14.0 | 9366 | 0.0017 | 0.9998 | 0.9999 |
| 0.0 | 15.0 | 10035 | 0.0000 | 1.0 | 1.0 |
| 0.0 | 16.0 | 10704 | 0.0000 | 1.0 | 1.0 |
| 0.0022 | 17.0 | 11373 | 0.0111 | 0.9981 | 0.9986 |
| 0.0004 | 18.0 | 12042 | 0.0011 | 0.9994 | 0.9996 |
| 0.0 | 19.0 | 12711 | 0.0020 | 0.9994 | 0.9996 |
| 0.0 | 20.0 | 13380 | 0.0023 | 0.9993 | 0.9995 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1