metadata
license: mit
base_model: facebook/esm2_t33_650M_UR50D
tags:
- generated_from_trainer
metrics:
- spearmanr
model-index:
- name: esm2_t33_650M_UR50D-finetuned-Ab14H-v1
results: []
esm2_t33_650M_UR50D-finetuned-Ab14H-v1
This model is a fine-tuned version of facebook/esm2_t33_650M_UR50D on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.2072
- Spearmanr: 0.6411
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: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 7
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Spearmanr |
|---|---|---|---|---|
| 3.3764 | 1.0 | 643 | 1.4308 | 0.6338 |
| 1.3787 | 2.0 | 1286 | 1.3095 | 0.6546 |
| 1.2298 | 3.0 | 1929 | 1.1653 | 0.6540 |
| 1.0901 | 4.0 | 2572 | 1.2472 | 0.6532 |
| 0.9758 | 5.0 | 3215 | 1.1286 | 0.6615 |
| 0.874 | 6.0 | 3858 | 1.1729 | 0.6485 |
| 0.6273 | 7.0 | 4501 | 1.2072 | 0.6411 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1