metadata
library_name: transformers
license: apache-2.0
base_model: PekingU/rtdetr_r50vd
tags:
- generated_from_trainer
model-index:
- name: rtdetr-cppe5-detection
results: []
rtdetr-cppe5-detection
This model is a fine-tuned version of PekingU/rtdetr_r50vd on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 89.3171
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: 1e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 237.8671 | 1.0 | 32 | 120.8588 |
| 160.6229 | 2.0 | 64 | 87.9866 |
| 123.3155 | 3.0 | 96 | 94.9977 |
| 116.224 | 4.0 | 128 | 94.1750 |
| 119.0437 | 5.0 | 160 | 96.6042 |
| 123.0258 | 6.0 | 192 | 95.5089 |
| 106.028 | 7.0 | 224 | 91.3961 |
| 104.9835 | 8.0 | 256 | 90.3529 |
| 103.8652 | 9.0 | 288 | 95.9664 |
| 107.2287 | 10.0 | 320 | 89.3171 |
Framework versions
- Transformers 4.57.3
- Pytorch 2.9.1+cu128
- Datasets 4.4.1
- Tokenizers 0.22.1