File size: 4,525 Bytes
f54f344 1e38cb4 f54f344 1e38cb4 f54f344 1e38cb4 f54f344 |
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 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
---
license: apache-2.0
tags:
- image-classification
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: new_exper3
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. -->
# new_exper3
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the sudo-s/herbier_mesuem1 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3000
- Accuracy: 0.9298
## 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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
- mixed_precision_training: Apex, opt level O1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 4.093 | 0.16 | 100 | 4.1045 | 0.1885 |
| 3.5057 | 0.31 | 200 | 3.4448 | 0.3231 |
| 2.9116 | 0.47 | 300 | 2.9483 | 0.4537 |
| 2.561 | 0.63 | 400 | 2.5700 | 0.5258 |
| 2.1611 | 0.78 | 500 | 2.1721 | 0.6145 |
| 1.715 | 0.94 | 600 | 1.8255 | 0.6407 |
| 1.2752 | 1.1 | 700 | 1.5340 | 0.7051 |
| 1.2487 | 1.25 | 800 | 1.3533 | 0.7201 |
| 1.0333 | 1.41 | 900 | 1.1474 | 0.7826 |
| 0.8856 | 1.56 | 1000 | 1.0914 | 0.7645 |
| 0.7512 | 1.72 | 1100 | 0.8893 | 0.8119 |
| 0.747 | 1.88 | 1200 | 0.8370 | 0.8304 |
| 0.5082 | 2.03 | 1300 | 0.7131 | 0.8566 |
| 0.4449 | 2.19 | 1400 | 0.6573 | 0.8547 |
| 0.2912 | 2.35 | 1500 | 0.6184 | 0.8597 |
| 0.285 | 2.5 | 1600 | 0.5974 | 0.8570 |
| 0.2267 | 2.66 | 1700 | 0.5621 | 0.8647 |
| 0.2553 | 2.82 | 1800 | 0.5044 | 0.8816 |
| 0.2029 | 2.97 | 1900 | 0.4342 | 0.8955 |
| 0.1763 | 3.13 | 2000 | 0.4487 | 0.8905 |
| 0.1418 | 3.29 | 2100 | 0.4173 | 0.9005 |
| 0.0563 | 3.44 | 2200 | 0.3870 | 0.9048 |
| 0.0579 | 3.6 | 2300 | 0.3849 | 0.9036 |
| 0.166 | 3.76 | 2400 | 0.3933 | 0.9025 |
| 0.11 | 3.91 | 2500 | 0.3918 | 0.9056 |
| 0.0356 | 4.07 | 2600 | 0.3298 | 0.9202 |
| 0.0513 | 4.23 | 2700 | 0.3371 | 0.9210 |
| 0.0762 | 4.38 | 2800 | 0.3253 | 0.9225 |
| 0.018 | 4.54 | 2900 | 0.3467 | 0.9148 |
| 0.0263 | 4.69 | 3000 | 0.3544 | 0.9144 |
| 0.0205 | 4.85 | 3100 | 0.3340 | 0.9221 |
| 0.0237 | 5.01 | 3200 | 0.3353 | 0.9144 |
| 0.013 | 5.16 | 3300 | 0.3218 | 0.9229 |
| 0.0116 | 5.32 | 3400 | 0.3088 | 0.9291 |
| 0.0119 | 5.48 | 3500 | 0.3047 | 0.9279 |
| 0.0098 | 5.63 | 3600 | 0.3063 | 0.9283 |
| 0.0086 | 5.79 | 3700 | 0.3074 | 0.9268 |
| 0.0081 | 5.95 | 3800 | 0.3220 | 0.9237 |
| 0.0078 | 6.1 | 3900 | 0.3064 | 0.9268 |
| 0.0074 | 6.26 | 4000 | 0.3062 | 0.9279 |
| 0.0068 | 6.42 | 4100 | 0.3051 | 0.9291 |
| 0.006 | 6.57 | 4200 | 0.3000 | 0.9298 |
| 0.0075 | 6.73 | 4300 | 0.3010 | 0.9310 |
| 0.0057 | 6.89 | 4400 | 0.3037 | 0.9298 |
| 0.0058 | 7.04 | 4500 | 0.3071 | 0.9279 |
| 0.0075 | 7.2 | 4600 | 0.3075 | 0.9283 |
| 0.0066 | 7.36 | 4700 | 0.3077 | 0.9295 |
| 0.0056 | 7.51 | 4800 | 0.3084 | 0.9295 |
| 0.0053 | 7.67 | 4900 | 0.3064 | 0.9310 |
| 0.0057 | 7.82 | 5000 | 0.3068 | 0.9318 |
| 0.0055 | 7.98 | 5100 | 0.3068 | 0.9318 |
### Framework versions
- Transformers 4.19.4
- Pytorch 1.5.1
- Datasets 2.3.2
- Tokenizers 0.12.1
|