metadata
license: apache-2.0
base_model: docketanalyzer/docket-lm-xs
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: label-complaint
results: []
label-complaint
This model is a fine-tuned version of docketanalyzer/docket-lm-xs on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0230
- F1: 0.9915
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 1
Training results
Training Loss | Epoch | Step | Validation Loss | F1 |
---|---|---|---|---|
0.0112 | 0.0418 | 300 | 0.0576 | 0.9771 |
0.0551 | 0.0836 | 600 | 0.0362 | 0.9857 |
0.2331 | 0.1254 | 900 | 0.0354 | 0.9839 |
0.0009 | 0.1672 | 1200 | 0.0396 | 0.9868 |
0.005 | 0.2090 | 1500 | 0.0526 | 0.9867 |
0.0948 | 0.2508 | 1800 | 0.0434 | 0.9865 |
0.016 | 0.2926 | 2100 | 0.0297 | 0.9876 |
0.0047 | 0.3344 | 2400 | 0.0394 | 0.9882 |
0.0007 | 0.3763 | 2700 | 0.0422 | 0.9864 |
0.0037 | 0.4181 | 3000 | 0.0248 | 0.9910 |
0.002 | 0.4599 | 3300 | 0.0271 | 0.9909 |
0.0005 | 0.5017 | 3600 | 0.0283 | 0.9902 |
0.0155 | 0.5435 | 3900 | 0.0227 | 0.9910 |
0.0017 | 0.5853 | 4200 | 0.0290 | 0.9907 |
0.0002 | 0.6271 | 4500 | 0.0264 | 0.9899 |
0.0051 | 0.6689 | 4800 | 0.0294 | 0.9907 |
0.0152 | 0.7107 | 5100 | 0.0253 | 0.9903 |
0.0096 | 0.7525 | 5400 | 0.0232 | 0.9909 |
0.1812 | 0.7943 | 5700 | 0.0295 | 0.9915 |
0.0007 | 0.8361 | 6000 | 0.0235 | 0.9912 |
0.0081 | 0.8779 | 6300 | 0.0247 | 0.9910 |
0.0684 | 0.9197 | 6600 | 0.0236 | 0.9905 |
0.0003 | 0.9615 | 6900 | 0.0230 | 0.9914 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.14.4
- Tokenizers 0.19.1