metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: mistralai/Mistral-7B-Instruct-v0.1
model-index:
- name: Mistral-7B-Instruct-v0_1-Txt-2-Presto-SQL
results: []
Mistral-7B-Instruct-v0_1-Txt-2-Presto-SQL
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.1 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.6492
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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_steps: 0.03
- training_steps: 80
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.3538 | 0.71 | 10 | 1.0781 |
| 1.0194 | 1.43 | 20 | 0.8769 |
| 0.8536 | 2.14 | 30 | 0.7753 |
| 0.7679 | 2.86 | 40 | 0.7234 |
| 0.7069 | 3.57 | 50 | 0.6825 |
| 0.6551 | 4.29 | 60 | 0.6609 |
| 0.6105 | 5.0 | 70 | 0.6500 |
| 0.582 | 5.71 | 80 | 0.6492 |
Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0