metadata
license: mit
base_model: FacebookAI/xlm-roberta-large
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: xlm-roberta-large-finetuned-ner-harem
results: []
xlm-roberta-large-finetuned-ner-harem
This model is a fine-tuned version of FacebookAI/xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1622
- Precision: 0.8344
- Recall: 0.8412
- F1: 0.8378
- Accuracy: 0.9745
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 0.9938 | 140 | 0.1806 | 0.6310 | 0.6557 | 0.6431 | 0.9533 |
| No log | 1.9947 | 281 | 0.1334 | 0.7314 | 0.7691 | 0.7497 | 0.9642 |
| No log | 2.9956 | 422 | 0.1332 | 0.7751 | 0.8103 | 0.7923 | 0.9712 |
| 0.2049 | 3.9965 | 563 | 0.1133 | 0.7948 | 0.8144 | 0.8045 | 0.9706 |
| 0.2049 | 4.9973 | 704 | 0.1215 | 0.814 | 0.8392 | 0.8264 | 0.9748 |
| 0.2049 | 5.9982 | 845 | 0.1274 | 0.8097 | 0.8247 | 0.8172 | 0.9726 |
| 0.2049 | 6.9991 | 986 | 0.1725 | 0.8079 | 0.8062 | 0.8070 | 0.9687 |
| 0.0307 | 8.0 | 1127 | 0.1647 | 0.8396 | 0.8309 | 0.8352 | 0.9736 |
| 0.0307 | 8.9938 | 1267 | 0.1678 | 0.8420 | 0.8351 | 0.8385 | 0.9726 |
| 0.0307 | 9.9379 | 1400 | 0.1622 | 0.8344 | 0.8412 | 0.8378 | 0.9745 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.1.2
- Datasets 2.19.1
- Tokenizers 0.19.1