metadata
license: apache-2.0
base_model: distilgpt2
tags:
- generated_from_trainer
model-index:
- name: distilgpt2-alpaca-instruction-fine-tuning-qlora
results: []
distilgpt2-alpaca-instruction-fine-tuning-qlora
This model is a fine-tuned version of distilgpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.7754
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: 0.0005
- train_batch_size: 24
- eval_batch_size: 24
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 1000
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.6372 | 0.66 | 1000 | 2.8398 |
| 3.0548 | 1.32 | 2000 | 2.7754 |
| 3.0124 | 1.99 | 3000 | 2.7832 |
| 3.0056 | 2.65 | 4000 | 2.8105 |
| 2.9911 | 3.31 | 5000 | 2.7812 |
| 2.983 | 3.97 | 6000 | 2.7676 |
| 2.9735 | 4.64 | 7000 | 2.7754 |
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3