metadata
base_model: zake7749/gemma-2-2b-it-chinese-kyara-dpo
library_name: peft
license: gemma
tags:
- generated_from_trainer
model-index:
- name: adl-hw3-finetune-gemma-2-chinese-kyara-5
results: []
adl-hw3-finetune-gemma-2-chinese-kyara-5
This model is a fine-tuned version of zake7749/gemma-2-2b-it-chinese-kyara-dpo on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.4426
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.3055 | 0.3556 | 200 | 2.1449 |
| 2.107 | 0.7111 | 400 | 2.0629 |
| 1.9753 | 1.0667 | 600 | 2.0500 |
| 1.6997 | 1.4222 | 800 | 2.0477 |
| 1.7306 | 1.7778 | 1000 | 2.0290 |
| 1.5365 | 2.1333 | 1200 | 2.1725 |
| 1.2898 | 2.4889 | 1400 | 2.1792 |
| 1.3049 | 2.8444 | 1600 | 2.1651 |
| 1.1339 | 3.2 | 1800 | 2.4387 |
| 0.9677 | 3.5556 | 2000 | 2.4493 |
| 0.9654 | 3.9111 | 2200 | 2.4426 |
Framework versions
- PEFT 0.13.2
- Transformers 4.45.1
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.20.2