Quick Links
dl_a3_q3_results
This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-ro on the wmt16 dataset. It achieves the following results on the evaluation set:
- Loss: 2.5716
- Bleu-1: 0.785
- Bleu-2: 0.6566
- Rouge-l: 0.3001
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu-1 | Bleu-2 | Rouge-l |
|---|---|---|---|---|---|---|
| 2.8855 | 1.0 | 1250 | 3.1121 | 0.7437 | 0.5967 | 0.2186 |
| 2.2274 | 2.0 | 2500 | 2.8313 | 0.749 | 0.6149 | 0.2604 |
| 2.019 | 3.0 | 3750 | 2.6962 | 0.7661 | 0.6359 | 0.2851 |
| 1.8669 | 4.0 | 5000 | 2.6241 | 0.7854 | 0.6544 | 0.2926 |
| 1.805 | 5.0 | 6250 | 2.5832 | 0.7812 | 0.6525 | 0.2987 |
| 1.7505 | 6.0 | 7500 | 2.5716 | 0.785 | 0.6566 | 0.3001 |
Framework versions
- Transformers 4.27.4
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.2
- Downloads last month
- 54
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("nks18/dl_a3_q3_results") model = AutoModelForSeq2SeqLM.from_pretrained("nks18/dl_a3_q3_results")