Quick Links
mt5-base
This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0516
- Wer: 0.0392
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: 0.0003
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 3.4396 | 0.81 | 4000 | 0.3368 | 0.3043 |
| 0.4257 | 1.62 | 8000 | 0.1328 | 0.1205 |
| 0.2185 | 2.42 | 12000 | 0.0929 | 0.0879 |
| 0.1506 | 3.23 | 16000 | 0.0762 | 0.0708 |
| 0.1133 | 4.04 | 20000 | 0.0663 | 0.0587 |
| 0.092 | 4.85 | 24000 | 0.0620 | 0.0551 |
| 0.0739 | 5.66 | 28000 | 0.0583 | 0.0507 |
| 0.0649 | 6.46 | 32000 | 0.0572 | 0.0465 |
| 0.0564 | 7.27 | 36000 | 0.0545 | 0.0439 |
| 0.0494 | 8.08 | 40000 | 0.0533 | 0.0425 |
| 0.0433 | 8.89 | 44000 | 0.0522 | 0.0405 |
| 0.0396 | 9.7 | 48000 | 0.0516 | 0.0392 |
Framework versions
- Transformers 4.33.0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3
- Downloads last month
- 13
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Ayon128/mt5-base
Base model
google/mt5-small
# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("Ayon128/mt5-base") model = AutoModelForSeq2SeqLM.from_pretrained("Ayon128/mt5-base")