|
--- |
|
license: apache-2.0 |
|
base_model: microsoft/swinv2-base-patch4-window8-256 |
|
tags: |
|
- image-classification |
|
- vision |
|
- generated_from_trainer |
|
datasets: |
|
- generator |
|
model-index: |
|
- name: swinv2-base-panorama-IQA |
|
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. --> |
|
|
|
# swinv2-base-panorama-IQA |
|
|
|
This model is a fine-tuned version of [microsoft/swinv2-base-patch4-window8-256](https://huggingface.co/microsoft/swinv2-base-patch4-window8-256) on the isiqa-2019-hf dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0312 |
|
- Srocc: 0.1132 |
|
- Lcc: 0.1583 |
|
|
|
## 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: 1e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 32 |
|
- seed: 10 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 64 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_ratio: 0.1 |
|
- num_epochs: 50.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Srocc | Lcc | |
|
|:-------------:|:-------:|:----:|:---------------:|:-------:|:-------:| |
|
| No log | 0.8571 | 3 | 0.3021 | -0.1668 | -0.1392 | |
|
| No log | 2.0 | 7 | 0.1286 | -0.1808 | -0.1347 | |
|
| 0.2494 | 2.8571 | 10 | 0.0678 | -0.1784 | -0.1273 | |
|
| 0.2494 | 4.0 | 14 | 0.1143 | -0.1625 | -0.1114 | |
|
| 0.2494 | 4.8571 | 17 | 0.0686 | -0.1939 | -0.1152 | |
|
| 0.069 | 6.0 | 21 | 0.0572 | -0.2063 | -0.1376 | |
|
| 0.069 | 6.8571 | 24 | 0.0537 | -0.1965 | -0.1405 | |
|
| 0.069 | 8.0 | 28 | 0.0671 | -0.1794 | -0.1289 | |
|
| 0.0276 | 8.8571 | 31 | 0.0551 | -0.1443 | -0.1164 | |
|
| 0.0276 | 10.0 | 35 | 0.0492 | -0.1110 | -0.0948 | |
|
| 0.0276 | 10.8571 | 38 | 0.0465 | -0.0945 | -0.0767 | |
|
| 0.0181 | 12.0 | 42 | 0.0449 | -0.0830 | -0.0464 | |
|
| 0.0181 | 12.8571 | 45 | 0.0402 | -0.0659 | -0.0280 | |
|
| 0.0181 | 14.0 | 49 | 0.0389 | -0.0411 | -0.0117 | |
|
| 0.0128 | 14.8571 | 52 | 0.0380 | -0.0348 | -0.0055 | |
|
| 0.0128 | 16.0 | 56 | 0.0371 | -0.0232 | 0.0088 | |
|
| 0.0128 | 16.8571 | 59 | 0.0360 | 0.0048 | 0.0205 | |
|
| 0.0112 | 18.0 | 63 | 0.0354 | 0.0128 | 0.0385 | |
|
| 0.0112 | 18.8571 | 66 | 0.0352 | 0.0197 | 0.0509 | |
|
| 0.0088 | 20.0 | 70 | 0.0346 | 0.0331 | 0.0670 | |
|
| 0.0088 | 20.8571 | 73 | 0.0337 | 0.0412 | 0.0801 | |
|
| 0.0088 | 22.0 | 77 | 0.0347 | 0.0396 | 0.0879 | |
|
| 0.008 | 22.8571 | 80 | 0.0348 | 0.0512 | 0.0954 | |
|
| 0.008 | 24.0 | 84 | 0.0339 | 0.0643 | 0.1071 | |
|
| 0.008 | 24.8571 | 87 | 0.0332 | 0.0765 | 0.1143 | |
|
| 0.0066 | 26.0 | 91 | 0.0334 | 0.0855 | 0.1240 | |
|
| 0.0066 | 26.8571 | 94 | 0.0330 | 0.0938 | 0.1292 | |
|
| 0.0066 | 28.0 | 98 | 0.0317 | 0.0997 | 0.1381 | |
|
| 0.006 | 28.8571 | 101 | 0.0314 | 0.1087 | 0.1432 | |
|
| 0.006 | 30.0 | 105 | 0.0317 | 0.1053 | 0.1446 | |
|
| 0.006 | 30.8571 | 108 | 0.0317 | 0.0971 | 0.1465 | |
|
| 0.0062 | 32.0 | 112 | 0.0315 | 0.1032 | 0.1496 | |
|
| 0.0062 | 32.8571 | 115 | 0.0315 | 0.1032 | 0.1511 | |
|
| 0.0062 | 34.0 | 119 | 0.0314 | 0.1032 | 0.1533 | |
|
| 0.0057 | 34.8571 | 122 | 0.0314 | 0.1094 | 0.1543 | |
|
| 0.0057 | 36.0 | 126 | 0.0313 | 0.1091 | 0.1558 | |
|
| 0.0057 | 36.8571 | 129 | 0.0312 | 0.1132 | 0.1570 | |
|
| 0.006 | 38.0 | 133 | 0.0312 | 0.1132 | 0.1577 | |
|
| 0.006 | 38.8571 | 136 | 0.0312 | 0.1132 | 0.1581 | |
|
| 0.0058 | 40.0 | 140 | 0.0312 | 0.1132 | 0.1583 | |
|
| 0.0058 | 40.8571 | 143 | 0.0312 | 0.1132 | 0.1584 | |
|
| 0.0058 | 42.0 | 147 | 0.0312 | 0.1132 | 0.1584 | |
|
| 0.006 | 42.8571 | 150 | 0.0312 | 0.1132 | 0.1584 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.42.3 |
|
- Pytorch 2.1.2 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
|
|
