metadata
license: mit
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: microsoft/phi-2
model-index:
- name: phi2-results2
results: []
phi2-results2
This model is a fine-tuned version of microsoft/phi-2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.3743
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 12
- total_train_batch_size: 12
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.3
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.2952 | 0.97 | 14 | 1.0035 |
| 1.0151 | 2.0 | 29 | 0.6051 |
| 0.5514 | 2.97 | 43 | 0.4484 |
| 0.4471 | 4.0 | 58 | 0.3983 |
| 0.4017 | 4.83 | 70 | 0.3743 |
Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0