metadata
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: cria-babylm2-subset-default-1e-3
results: []
cria-babylm2-subset-default-1e-3
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.6626
- Accuracy: 0.5204
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: 0.001
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 32000
- num_epochs: 10.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.4397 | 1.0 | 14142 | 2.6684 | 0.4955 |
| 2.3085 | 2.0 | 28284 | 2.5420 | 0.5093 |
| 2.19 | 3.0 | 42426 | 2.4397 | 0.5215 |
| 2.0865 | 4.0 | 56568 | 2.3943 | 0.5276 |
| 1.9957 | 5.0 | 70710 | 2.3786 | 0.5305 |
| 1.9161 | 6.0 | 84852 | 2.3910 | 0.5313 |
| 1.8361 | 7.0 | 98994 | 2.4205 | 0.5304 |
| 1.7477 | 8.0 | 113136 | 2.4748 | 0.5283 |
| 1.6549 | 9.0 | 127278 | 2.5582 | 0.5249 |
| 1.5611 | 10.0 | 141420 | 2.6626 | 0.5204 |
Framework versions
- Transformers 4.42.4
- Pytorch 2.2.0+cu121
- Datasets 2.16.1
- Tokenizers 0.19.1