QAGP_final
This model is a fine-tuned version of google/pegasus-xsum on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2268
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: 0.0001
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.8819 | 0.16 | 1000 | 1.4078 |
| 1.4792 | 0.33 | 2000 | 1.3247 |
| 1.4406 | 0.49 | 3000 | 1.2776 |
| 1.4286 | 0.66 | 4000 | 1.2464 |
| 1.3776 | 0.82 | 5000 | 1.2476 |
| 1.3361 | 0.98 | 6000 | 1.2191 |
| 1.0983 | 1.15 | 7000 | 1.2404 |
| 1.0992 | 1.31 | 8000 | 1.2255 |
| 1.064 | 1.48 | 9000 | 1.2169 |
| 1.0771 | 1.64 | 10000 | 1.2077 |
| 1.0484 | 1.8 | 11000 | 1.2057 |
| 1.0428 | 1.97 | 12000 | 1.1912 |
| 0.8319 | 2.13 | 13000 | 1.2383 |
| 0.8801 | 2.3 | 14000 | 1.2271 |
| 0.8215 | 2.46 | 15000 | 1.2353 |
| 0.8407 | 2.63 | 16000 | 1.2328 |
| 0.8449 | 2.79 | 17000 | 1.2258 |
| 0.8283 | 2.95 | 18000 | 1.2268 |
Framework versions
- Transformers 4.22.2
- Pytorch 1.12.1+cu113
- Tokenizers 0.12.1