byt5-base-1e4-jit-tagged-7epochs-jamoparams-a100-chunked-bs8-ga16-eb128
This model is a fine-tuned version of google/byt5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.0111
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: 0.0005
- train_batch_size: 8
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Use adamw_torch with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: inverse_sqrt
- lr_scheduler_warmup_steps: 4000
- num_epochs: 7
- label_smoothing_factor: 0.1
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.1145 | 1.0 | 2507 | 1.0737 |
| 1.055 | 2.0 | 5014 | 1.0367 |
| 1.0322 | 3.0 | 7521 | 1.0224 |
| 1.0187 | 4.0 | 10028 | 1.0160 |
| 1.0097 | 5.0 | 12535 | 1.0129 |
| 1.0021 | 6.0 | 15042 | 1.0108 |
| 0.9956 | 7.0 | 17549 | 1.0111 |
Framework versions
- Transformers 4.57.3
- Pytorch 2.9.1+cu128
- Datasets 4.4.1
- Tokenizers 0.22.1
- Downloads last month
- 5
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for jinx2321/byt5-base-1e4-jit-tagged-7epochs-jamoparams-a100-chunked-bs8-ga16-eb128
Base model
google/byt5-base