metadata
license: apache-2.0
base_model: tiiuae/falcon-7b
tags:
- generated_from_trainer
datasets:
- spider
model-index:
- name: falcon_qlora_sql
results: []
falcon_qlora_sql
This model is a fine-tuned version of tiiuae/falcon-7b on the spider dataset. It achieves the following results on the evaluation set:
- Loss: 0.1772
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 43.7
- num_epochs: 3
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
0.2917 | 0.23 | 100 | 0.3571 |
0.8787 | 0.46 | 200 | 0.3685 |
0.3366 | 0.68 | 300 | 0.4001 |
0.2319 | 0.91 | 400 | 0.3134 |
0.3005 | 1.14 | 500 | 0.3567 |
0.4231 | 1.37 | 600 | 0.3052 |
0.2535 | 1.6 | 700 | 0.3218 |
0.2257 | 1.83 | 800 | 0.2744 |
0.0771 | 2.05 | 900 | 0.2567 |
0.0978 | 2.28 | 1000 | 0.2304 |
0.2031 | 2.51 | 1100 | 0.2236 |
0.2471 | 2.74 | 1200 | 0.2431 |
0.1514 | 2.97 | 1300 | 0.1897 |
Framework versions
- Transformers 4.31.0.dev0
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3