metadata
library_name: peft
license: mit
base_model: microsoft/Phi-3.5-mini-instruct
tags:
- generated_from_trainer
datasets:
- scitldr
model-index:
- name: Phi-3.5-Mini-Instruct-Summarization-QLoRa
results: []
pipeline_tag: summarization
Phi-3.5-Mini-Instruct-Summarization-QLoRa
This model is a fine-tuned version of microsoft/Phi-3.5-mini-instruct on the scitldr dataset. It achieves the following results on the evaluation set:
- Loss: 2.1376
Model description
More information needed
Intended uses & limitations
Summarization
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
- optimizer: Use paged_adamw_32bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.0519 | 0.2510 | 500 | 2.1280 |
| 2.0279 | 0.5020 | 1000 | 2.1223 |
| 2.0514 | 0.7530 | 1500 | 2.1131 |
| 2.0313 | 1.0040 | 2000 | 2.1142 |
| 1.8923 | 1.2550 | 2500 | 2.1390 |
| 1.8487 | 1.5060 | 3000 | 2.1375 |
| 1.819 | 1.7570 | 3500 | 2.1376 |
Framework versions
- PEFT 0.14.0
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0