File size: 4,081 Bytes
d8c5c1b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 | ---
library_name: transformers
license: apache-2.0
base_model: timm/tf_efficientnetv2_s.in21k
tags:
- timm
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: EfficientNetV2_Small_v1
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. -->
# EfficientNetV2_Small_v1
This model is a fine-tuned version of [timm/tf_efficientnetv2_s.in21k](https://huggingface.co/timm/tf_efficientnetv2_s.in21k) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0340
- Accuracy: 0.9935
- Precision: 0.9981
- Recall: 0.9878
- F1: 0.9929
- Tp: 1618
- Tn: 1907
- Fp: 3
- Fn: 20
## 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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 442
- num_epochs: 20
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Tp | Tn | Fp | Fn |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:----:|:----:|:--:|:--:|
| 0.1995 | 1.0 | 222 | 0.1349 | 0.9628 | 0.9575 | 0.9621 | 0.9598 | 1576 | 1840 | 70 | 62 |
| 0.1442 | 2.0 | 444 | 0.0940 | 0.9789 | 0.9956 | 0.9585 | 0.9767 | 1570 | 1903 | 7 | 68 |
| 0.1625 | 3.0 | 666 | 0.0827 | 0.9837 | 0.9925 | 0.9719 | 0.9821 | 1592 | 1898 | 12 | 46 |
| 0.1592 | 4.0 | 888 | 0.0926 | 0.9752 | 0.9708 | 0.9756 | 0.9732 | 1598 | 1862 | 48 | 40 |
| 0.1100 | 5.0 | 1110 | 0.0544 | 0.9876 | 0.9950 | 0.9780 | 0.9865 | 1602 | 1902 | 8 | 36 |
| 0.1497 | 6.0 | 1332 | 0.0635 | 0.9868 | 0.9877 | 0.9835 | 0.9856 | 1611 | 1890 | 20 | 27 |
| 0.1125 | 7.0 | 1554 | 0.0485 | 0.9896 | 0.9957 | 0.9817 | 0.9886 | 1608 | 1903 | 7 | 30 |
| 0.1202 | 8.0 | 1776 | 0.0774 | 0.9794 | 0.9740 | 0.9817 | 0.9778 | 1608 | 1867 | 43 | 30 |
| 0.1031 | 9.0 | 1998 | 0.0507 | 0.9893 | 0.9938 | 0.9829 | 0.9883 | 1610 | 1900 | 10 | 28 |
| 0.1211 | 10.0 | 2220 | 0.0434 | 0.9915 | 0.9975 | 0.9841 | 0.9908 | 1612 | 1906 | 4 | 26 |
| 0.1239 | 11.0 | 2442 | 0.0400 | 0.9918 | 0.9975 | 0.9847 | 0.9911 | 1613 | 1906 | 4 | 25 |
| 0.1066 | 12.0 | 2664 | 0.0403 | 0.9927 | 0.9988 | 0.9853 | 0.9920 | 1614 | 1908 | 2 | 24 |
| 0.1065 | 13.0 | 2886 | 0.0363 | 0.9927 | 0.9994 | 0.9847 | 0.9920 | 1613 | 1909 | 1 | 25 |
| 0.1074 | 14.0 | 3108 | 0.0378 | 0.9930 | 0.9988 | 0.9860 | 0.9923 | 1615 | 1908 | 2 | 23 |
| 0.1128 | 15.0 | 3330 | 0.0327 | 0.9924 | 0.9981 | 0.9853 | 0.9917 | 1614 | 1907 | 3 | 24 |
| 0.0963 | 16.0 | 3552 | 0.0309 | 0.9930 | 0.9988 | 0.9860 | 0.9923 | 1615 | 1908 | 2 | 23 |
| 0.1379 | 17.0 | 3774 | 0.0366 | 0.9927 | 0.9969 | 0.9872 | 0.9920 | 1617 | 1905 | 5 | 21 |
| 0.1070 | 18.0 | 3996 | 0.0331 | 0.9930 | 0.9981 | 0.9866 | 0.9923 | 1616 | 1907 | 3 | 22 |
| 0.1332 | 19.0 | 4218 | 0.0343 | 0.9930 | 0.9981 | 0.9866 | 0.9923 | 1616 | 1907 | 3 | 22 |
| 0.1294 | 20.0 | 4440 | 0.0340 | 0.9935 | 0.9981 | 0.9878 | 0.9929 | 1618 | 1907 | 3 | 20 |
### Framework versions
- Transformers 5.2.0
- Pytorch 2.9.0+cu126
- Datasets 4.0.0
- Tokenizers 0.22.2
|