metadata
library_name: transformers
license: apache-2.0
base_model: facebook/dinov2-large
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: dinov2-large-5-epochs-finetune-linear-probe-trueface
results: []
dinov2-large-5-epochs-finetune-linear-probe-trueface
This model is a fine-tuned version of facebook/dinov2-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0943
- Accuracy: 0.9769
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: 5e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 4
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3506 | 0.9993 | 369 | 0.2202 | 0.9277 |
| 0.2614 | 1.9986 | 738 | 0.1336 | 0.9639 |
| 0.2325 | 2.9980 | 1107 | 0.1082 | 0.9731 |
| 0.214 | 4.0 | 1477 | 0.0979 | 0.9756 |
| 0.2142 | 4.9966 | 1845 | 0.0943 | 0.9769 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.0