metadata
license: mit
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: microsoft/phi-2
model-index:
- name: phi_2_dn
results: []
phi_2_dn
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: 0.9427
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: 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
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.135 | 1.0 | 43 | 1.1428 |
| 1.0211 | 2.0 | 86 | 1.0620 |
| 0.9549 | 3.0 | 129 | 1.0160 |
| 0.9086 | 4.0 | 172 | 0.9812 |
| 0.896 | 5.0 | 215 | 0.9631 |
| 0.8564 | 6.0 | 258 | 0.9427 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2