metadata
base_model: google/paligemma-3b-pt-224
library_name: peft
license: gemma
tags:
- generated_from_trainer
model-index:
- name: paligemma_scienceQA_v15
results: []
paligemma_scienceQA_v15
This model is a fine-tuned version of google/paligemma-3b-pt-224 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1043
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-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1502 | 0.2 | 10 | 0.1155 |
| 0.1414 | 0.4 | 20 | 0.1136 |
| 0.1481 | 0.6 | 30 | 0.1134 |
| 0.1384 | 0.8 | 40 | 0.1121 |
| 0.1472 | 1.0 | 50 | 0.1112 |
| 0.1357 | 1.2 | 60 | 0.1108 |
| 0.124 | 1.4 | 70 | 0.1102 |
| 0.1495 | 1.6 | 80 | 0.1096 |
| 0.1438 | 1.8 | 90 | 0.1090 |
| 0.1394 | 2.0 | 100 | 0.1078 |
| 0.1285 | 2.2 | 110 | 0.1071 |
| 0.131 | 2.4 | 120 | 0.1071 |
| 0.1375 | 2.6 | 130 | 0.1068 |
| 0.1243 | 2.8 | 140 | 0.1061 |
| 0.138 | 3.0 | 150 | 0.1056 |
| 0.1386 | 3.2 | 160 | 0.1055 |
| 0.1325 | 3.4 | 170 | 0.1052 |
| 0.1252 | 3.6 | 180 | 0.1052 |
| 0.1371 | 3.8 | 190 | 0.1049 |
| 0.122 | 4.0 | 200 | 0.1048 |
| 0.1149 | 4.2 | 210 | 0.1046 |
| 0.1318 | 4.4 | 220 | 0.1044 |
| 0.1169 | 4.6 | 230 | 0.1043 |
| 0.1412 | 4.8 | 240 | 0.1041 |
| 0.1514 | 5.0 | 250 | 0.1043 |
Framework versions
- PEFT 0.12.0
- Transformers 4.43.3
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1