metadata
license: apache-2.0
base_model: hfl/chinese-xlnet-mid
tags:
- generated_from_trainer
model-index:
- name: xlnet-mid
results: []
xlnet-mid
This model is a fine-tuned version of hfl/chinese-xlnet-mid on the None dataset. It achieves the following results on the evaluation set:
- Loss: 4.7725
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 500
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 5.623 | 0.08 | 500 | 5.6406 |
| 5.456 | 0.16 | 1000 | 5.5080 |
| 5.2287 | 0.23 | 1500 | 5.3861 |
| 4.9929 | 0.31 | 2000 | 5.3493 |
| 5.3415 | 0.39 | 2500 | 5.1613 |
| 5.2163 | 0.47 | 3000 | 5.0551 |
| 5.1069 | 0.55 | 3500 | 4.9676 |
| 5.0365 | 0.62 | 4000 | 4.9118 |
| 4.9674 | 0.7 | 4500 | 4.8411 |
| 4.9259 | 0.78 | 5000 | 4.7963 |
| 4.8875 | 0.86 | 5500 | 4.7777 |
| 4.9003 | 0.94 | 6000 | 4.7725 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.0
- Tokenizers 0.15.0