Downloads · 30 days
6
1% of all-time downloads
SwinMark2
Abhiram4/SwinMark2
SwinMark2 is a image classification model from Abhiram4. Use it when you need a label for an image. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
416
Public
Repo size
442 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin110 MB · 100%
At a glance
- Task
- Image Classification
- Library
- transformers
- License
- apache-2.0
- Model type
- swinv2
- Access
- Public
- Created
- Oct 18, 2023
- Updated
- Nov 21, 2023
- SHA
- 2e73fba0
Base models
Datasets
Model card
From the Hugging Face model README
<!-- 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. -->
SwinMark2
This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the image_folder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0952
- Accuracy: 0.9666
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: 5e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.1407 | 1.0 | 231 | 0.1230 | 0.9586 |
| 0.1209 | 2.0 | 462 | 0.1066 | 0.9630 |
| 0.0987 | 3.0 | 693 | 0.0952 | 0.9666 |
Framework versions
- Transformers 4.33.0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3
- Task
- Image Classification
- Library
- transformers
- Type
- swinv2
- License
- apache-2.0
- Created
- Oct 18, 2023
- Updated
- Nov 21, 2023