---
base_model: google/paligemma-3b-pt-224
library_name: peft
license: gemma
tags:
- generated_from_trainer
model-index:
- name: palige_original_lora_8_epo_12
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# palige_original_lora_8_epo_12

This model is a fine-tuned version of [google/paligemma-3b-pt-224](https://huggingface.co/google/paligemma-3b-pt-224) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8624

## 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: 2e-05
- train_batch_size: 10
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 40
- optimizer: Use OptimizerNames.ADAMW_HF with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- num_epochs: 12

### Training results

| Training Loss | Epoch   | Step | Validation Loss |
|:-------------:|:-------:|:----:|:---------------:|
| 4.1762        | 0.3125  | 100  | 2.4304          |
| 2.2687        | 0.625   | 200  | 1.5980          |
| 1.7935        | 0.9375  | 300  | 1.3086          |
| 1.6099        | 1.25    | 400  | 1.1783          |
| 1.5425        | 1.5625  | 500  | 1.0963          |
| 1.4383        | 1.875   | 600  | 1.0387          |
| 1.3341        | 2.1875  | 700  | 0.9937          |
| 1.319         | 2.5     | 800  | 0.9610          |
| 1.2648        | 2.8125  | 900  | 0.9275          |
| 1.0991        | 3.125   | 1000 | 0.9046          |
| 1.1596        | 3.4375  | 1100 | 0.8902          |
| 1.1586        | 3.75    | 1200 | 0.8788          |
| 1.0607        | 4.0625  | 1300 | 0.8631          |
| 1.0172        | 4.375   | 1400 | 0.8527          |
| 1.014         | 4.6875  | 1500 | 0.8435          |
| 1.004         | 5.0     | 1600 | 0.8319          |
| 0.8944        | 5.3125  | 1700 | 0.8404          |
| 0.9064        | 5.625   | 1800 | 0.8148          |
| 0.9136        | 5.9375  | 1900 | 0.8178          |
| 0.8312        | 6.25    | 2000 | 0.8241          |
| 0.8253        | 6.5625  | 2100 | 0.8118          |
| 0.8623        | 6.875   | 2200 | 0.8004          |
| 0.7545        | 7.1875  | 2300 | 0.8054          |
| 0.7508        | 7.5     | 2400 | 0.8135          |
| 0.7204        | 7.8125  | 2500 | 0.7960          |
| 0.6948        | 8.125   | 2600 | 0.8121          |
| 0.6607        | 8.4375  | 2700 | 0.8265          |
| 0.6804        | 8.75    | 2800 | 0.8077          |
| 0.6556        | 9.0625  | 2900 | 0.8440          |
| 0.5798        | 9.375   | 3000 | 0.8263          |
| 0.5787        | 9.6875  | 3100 | 0.8274          |
| 0.5909        | 10.0    | 3200 | 0.8241          |
| 0.5254        | 10.3125 | 3300 | 0.8548          |
| 0.5184        | 10.625  | 3400 | 0.8649          |
| 0.5267        | 10.9375 | 3500 | 0.8467          |
| 0.4687        | 11.25   | 3600 | 0.8830          |
| 0.4678        | 11.5625 | 3700 | 0.8639          |
| 0.4528        | 11.875  | 3800 | 0.8624          |


### Framework versions

- PEFT 0.13.0
- Transformers 4.46.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.0