---
base_model: google/paligemma-3b-pt-224
library_name: peft
license: gemma
tags:
- generated_from_trainer
model-index:
- name: paligemma_vqav2_lr1e_5
  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. -->

# paligemma_vqav2_lr1e_5

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: 0.8757

## 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 |
|:-------------:|:-------:|:----:|:---------------:|
| 4.8047        | 0.3125  | 100  | 2.7102          |
| 2.5191        | 0.625   | 200  | 1.8346          |
| 1.9491        | 0.9375  | 300  | 1.4737          |
| 1.7149        | 1.25    | 400  | 1.2760          |
| 1.638         | 1.5625  | 500  | 1.1683          |
| 1.5128        | 1.875   | 600  | 1.1024          |
| 1.4079        | 2.1875  | 700  | 1.0432          |
| 1.3928        | 2.5     | 800  | 1.0078          |
| 1.3355        | 2.8125  | 900  | 0.9682          |
| 1.1662        | 3.125   | 1000 | 0.9412          |
| 1.222         | 3.4375  | 1100 | 0.9246          |
| 1.2254        | 3.75    | 1200 | 0.9102          |
| 1.1127        | 4.0625  | 1300 | 0.8936          |
| 1.0704        | 4.375   | 1400 | 0.8698          |
| 1.0614        | 4.6875  | 1500 | 0.8617          |
| 1.0572        | 5.0     | 1600 | 0.8531          |
| 0.9373        | 5.3125  | 1700 | 0.8505          |
| 0.9517        | 5.625   | 1800 | 0.8339          |
| 0.9599        | 5.9375  | 1900 | 0.8308          |
| 0.8669        | 6.25    | 2000 | 0.8364          |
| 0.8671        | 6.5625  | 2100 | 0.8374          |
| 0.8996        | 6.875   | 2200 | 0.8258          |
| 0.7827        | 7.1875  | 2300 | 0.8214          |
| 0.7874        | 7.5     | 2400 | 0.8343          |
| 0.7524        | 7.8125  | 2500 | 0.8135          |
| 0.7243        | 8.125   | 2600 | 0.8300          |
| 0.6922        | 8.4375  | 2700 | 0.8323          |
| 0.7111        | 8.75    | 2800 | 0.8259          |
| 0.6894        | 9.0625  | 2900 | 0.8595          |
| 0.6122        | 9.375   | 3000 | 0.8308          |
| 0.6113        | 9.6875  | 3100 | 0.8358          |
| 0.6199        | 10.0    | 3200 | 0.8371          |
| 0.5587        | 10.3125 | 3300 | 0.8657          |
| 0.5487        | 10.625  | 3400 | 0.8677          |
| 0.5529        | 10.9375 | 3500 | 0.8612          |
| 0.4998        | 11.25   | 3600 | 0.8937          |
| 0.4999        | 11.5625 | 3700 | 0.8784          |
| 0.4809        | 11.875  | 3800 | 0.8757          |


### Framework versions

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