metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: MarianMix_en-10
results: []
MarianMix_en-10
This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-zh on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.0668
- Bleu: 17.4306
- Gen Len: 46.8231
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: 32
- eval_batch_size: 32
- seed: 99
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 1.9405 | 0.44 | 500 | 2.2660 | 0.0253 | 22.8633 |
| 1.1843 | 0.89 | 1000 | 2.0616 | 0.0018 | 21.1749 |
| 0.9938 | 1.33 | 1500 | 1.6499 | 2.3884 | 42.8784 |
| 0.8307 | 1.78 | 2000 | 1.4181 | 6.5182 | 43.9226 |
| 0.7281 | 2.22 | 2500 | 1.2915 | 9.27 | 44.205 |
| 0.6684 | 2.67 | 3000 | 1.2056 | 12.5594 | 45.7598 |
| 0.6297 | 3.11 | 3500 | 1.1529 | 14.6009 | 48.5106 |
| 0.5787 | 3.56 | 4000 | 1.1187 | 16.3988 | 47.403 |
| 0.5657 | 4.0 | 4500 | 1.0890 | 16.0136 | 45.2925 |
| 0.5305 | 4.45 | 5000 | 1.0750 | 17.1562 | 47.5196 |
| 0.5234 | 4.89 | 5500 | 1.0668 | 17.4306 | 46.8231 |
Framework versions
- Transformers 4.12.5
- Pytorch 1.9.1
- Datasets 1.17.0
- Tokenizers 0.10.3