Downloads · 30 days
84
3% of all-time downloads
legal-german-roberta-base
joelniklaus/legal-german-roberta-base
legal-german-roberta-base is a fill-mask model from joelniklaus. Use it when you need the model to fill a missing word. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
2.6K
Public
Repo size
27 GB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.bin221 MB · 99%
At a glance
- Task
- Fill-Mask
- Library
- transformers
- Model type
- roberta
- Access
- Public
- Created
- Dec 27, 2022
- Updated
- Jan 7, 2023
- SHA
- fb55c64d
Try a prompt
Model card
From the Hugging Face model README
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
legal-german-roberta-base
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.7080
- Accuracy: 0.8387
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: 0.0001
- train_batch_size: 1024
- eval_batch_size: 512
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.05
- training_steps: 1000000
Training results
| Training Loss | Epoch | Step | Accuracy | Validation Loss |
|---|---|---|---|---|
| 2.1008 | 0.05 | 50000 | 0.6533 | 2.0523 |
| 1.5248 | 0.1 | 100000 | 0.7661 | 1.1575 |
| 1.3152 | 0.15 | 150000 | 0.7674 | 1.1281 |
| 1.1239 | 0.2 | 200000 | 0.7971 | 0.9458 |
| 0.9472 | 0.25 | 250000 | 0.7876 | 0.9979 |
| 0.961 | 0.3 | 300000 | 0.8075 | 0.8798 |
| 1.0179 | 0.35 | 350000 | 0.8018 | 0.9102 |
| 1.037 | 0.4 | 400000 | 0.8195 | 0.8107 |
| 1.1206 | 0.45 | 450000 | 0.8152 | 0.8323 |
| 1.0865 | 0.5 | 500000 | 0.8242 | 0.7829 |
| 0.9616 | 0.55 | 550000 | 0.8224 | 0.7895 |
| 0.7727 | 0.6 | 600000 | 0.8285 | 0.7585 |
| 0.9871 | 1.04 | 650000 | 0.8320 | 0.7391 |
| 1.0679 | 1.09 | 700000 | 0.8311 | 0.7436 |
| 0.9203 | 1.14 | 750000 | 0.8355 | 0.7187 |
| 0.9626 | 1.19 | 800000 | 0.8353 | 0.7242 |
| 0.7263 | 1.24 | 850000 | 0.7094 | 0.8378 |
| 0.8578 | 1.29 | 900000 | 0.7140 | 0.8368 |
| 0.7693 | 1.34 | 950000 | 0.7091 | 0.8377 |
| 1.0488 | 1.39 | 1000000 | 0.7080 | 0.8387 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.10.0+cu113
- Datasets 2.8.0
- Tokenizers 0.12.1
- Task
- Fill-Mask
- Library
- transformers
- Type
- roberta
- Created
- Dec 27, 2022
- Updated
- Jan 7, 2023