Extractor_Adaptor_Qwen3_QA_websrc_lost_hope
This model is a fine-tuned version of Qwen/Qwen3-0.6B on the web_finetune_train dataset. It achieves the following results on the evaluation set:
- Loss: 0.0964
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: 8
- eval_batch_size: 10
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 1.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.0802 | 0.1067 | 50 | 0.1069 |
| 0.0405 | 0.2133 | 100 | 0.1138 |
| 0.0456 | 0.32 | 150 | 0.0964 |
| 0.0441 | 0.4267 | 200 | 0.1080 |
| 0.0268 | 0.5333 | 250 | 0.1100 |
| 0.0352 | 0.64 | 300 | 0.1110 |
| 0.0466 | 0.7467 | 350 | 0.1162 |
| 0.0308 | 0.8533 | 400 | 0.1127 |
| 0.0315 | 0.96 | 450 | 0.1133 |
Framework versions
- PEFT 0.17.1
- Transformers 4.57.1
- Pytorch 2.4.1+cu124
- Datasets 4.0.0
- Tokenizers 0.22.1
- Downloads last month
- -