metadata
license: mit
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: microsoft/Phi-3-medium-128k-instruct
model-index:
- name: results_medium
results: []
results_medium
This model is a fine-tuned version of microsoft/Phi-3-medium-128k-instruct on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1562
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.9442 | 0.2131 | 100 | 1.9833 |
| 1.8439 | 0.4262 | 200 | 1.8431 |
| 1.5387 | 0.6393 | 300 | 1.5840 |
| 1.4157 | 0.8524 | 400 | 1.3748 |
| 1.2628 | 1.0655 | 500 | 1.2339 |
| 1.2704 | 1.2786 | 600 | 1.1894 |
| 1.1634 | 1.4917 | 700 | 1.1763 |
| 1.2318 | 1.7048 | 800 | 1.1708 |
| 1.0665 | 1.9180 | 900 | 1.1683 |
| 1.1715 | 2.1311 | 1000 | 1.1634 |
| 1.1366 | 2.3442 | 1100 | 1.1617 |
| 1.1239 | 2.5573 | 1200 | 1.1584 |
| 1.1442 | 2.7704 | 1300 | 1.1568 |
| 1.1336 | 2.9835 | 1400 | 1.1562 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1