Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: facebook/dino-vitb8
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
+
- f1
|
| 10 |
+
- precision
|
| 11 |
+
- recall
|
| 12 |
+
model-index:
|
| 13 |
+
- name: dino-vitb8-finetuned-stroke-binary
|
| 14 |
+
results: []
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 18 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
+
|
| 20 |
+
# dino-vitb8-finetuned-stroke-binary
|
| 21 |
+
|
| 22 |
+
This model is a fine-tuned version of [facebook/dino-vitb8](https://huggingface.co/facebook/dino-vitb8) on an unknown dataset.
|
| 23 |
+
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.1127
|
| 25 |
+
- Accuracy: 0.9597
|
| 26 |
+
- F1: 0.9595
|
| 27 |
+
- Precision: 0.9602
|
| 28 |
+
- Recall: 0.9597
|
| 29 |
+
|
| 30 |
+
## Model description
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Intended uses & limitations
|
| 35 |
+
|
| 36 |
+
More information needed
|
| 37 |
+
|
| 38 |
+
## Training and evaluation data
|
| 39 |
+
|
| 40 |
+
More information needed
|
| 41 |
+
|
| 42 |
+
## Training procedure
|
| 43 |
+
|
| 44 |
+
### Training hyperparameters
|
| 45 |
+
|
| 46 |
+
The following hyperparameters were used during training:
|
| 47 |
+
- learning_rate: 2e-05
|
| 48 |
+
- train_batch_size: 8
|
| 49 |
+
- eval_batch_size: 8
|
| 50 |
+
- seed: 42
|
| 51 |
+
- gradient_accumulation_steps: 4
|
| 52 |
+
- total_train_batch_size: 32
|
| 53 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 54 |
+
- lr_scheduler_type: cosine
|
| 55 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 56 |
+
- num_epochs: 36
|
| 57 |
+
- mixed_precision_training: Native AMP
|
| 58 |
+
|
| 59 |
+
### Training results
|
| 60 |
+
|
| 61 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|
| 62 |
+
|:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
|
| 63 |
+
| 0.7965 | 0.6202 | 100 | 0.8312 | 0.5382 | 0.5058 | 0.4968 | 0.5382 |
|
| 64 |
+
| 0.6839 | 1.2357 | 200 | 0.6796 | 0.6246 | 0.5750 | 0.5991 | 0.6246 |
|
| 65 |
+
| 0.5662 | 1.8558 | 300 | 0.5344 | 0.7318 | 0.7119 | 0.7377 | 0.7318 |
|
| 66 |
+
| 0.4408 | 2.4713 | 400 | 0.4082 | 0.8123 | 0.8082 | 0.8120 | 0.8123 |
|
| 67 |
+
| 0.3611 | 3.0868 | 500 | 0.3335 | 0.8602 | 0.8597 | 0.8596 | 0.8602 |
|
| 68 |
+
| 0.3121 | 3.7070 | 600 | 0.2746 | 0.8860 | 0.8832 | 0.8914 | 0.8860 |
|
| 69 |
+
| 0.2614 | 4.3225 | 700 | 0.2299 | 0.9050 | 0.9040 | 0.9058 | 0.9050 |
|
| 70 |
+
| 0.242 | 4.9426 | 800 | 0.2103 | 0.9177 | 0.9178 | 0.9179 | 0.9177 |
|
| 71 |
+
| 0.2239 | 5.5581 | 900 | 0.2298 | 0.9082 | 0.9090 | 0.9136 | 0.9082 |
|
| 72 |
+
| 0.1979 | 6.1736 | 1000 | 0.2059 | 0.9209 | 0.9197 | 0.9230 | 0.9209 |
|
| 73 |
+
| 0.2082 | 6.7938 | 1100 | 0.1779 | 0.9263 | 0.9261 | 0.9261 | 0.9263 |
|
| 74 |
+
| 0.1723 | 7.4093 | 1200 | 0.1693 | 0.9308 | 0.9302 | 0.9315 | 0.9308 |
|
| 75 |
+
| 0.1877 | 8.0248 | 1300 | 0.1681 | 0.9380 | 0.9382 | 0.9385 | 0.9380 |
|
| 76 |
+
| 0.2 | 8.6450 | 1400 | 0.1482 | 0.9403 | 0.9402 | 0.9402 | 0.9403 |
|
| 77 |
+
| 0.1642 | 9.2605 | 1500 | 0.1637 | 0.9331 | 0.9322 | 0.9352 | 0.9331 |
|
| 78 |
+
| 0.1525 | 9.8806 | 1600 | 0.1494 | 0.9421 | 0.9417 | 0.9425 | 0.9421 |
|
| 79 |
+
| 0.158 | 10.4961 | 1700 | 0.1403 | 0.9484 | 0.9480 | 0.9495 | 0.9484 |
|
| 80 |
+
| 0.1327 | 11.1116 | 1800 | 0.1329 | 0.9498 | 0.9498 | 0.9498 | 0.9498 |
|
| 81 |
+
| 0.1465 | 11.7318 | 1900 | 0.1233 | 0.9525 | 0.9524 | 0.9525 | 0.9525 |
|
| 82 |
+
| 0.1311 | 12.3473 | 2000 | 0.1280 | 0.9521 | 0.9520 | 0.9520 | 0.9521 |
|
| 83 |
+
| 0.129 | 12.9674 | 2100 | 0.1173 | 0.9557 | 0.9556 | 0.9556 | 0.9557 |
|
| 84 |
+
| 0.1425 | 13.5829 | 2200 | 0.1190 | 0.9552 | 0.9552 | 0.9552 | 0.9552 |
|
| 85 |
+
| 0.1256 | 14.1984 | 2300 | 0.1225 | 0.9566 | 0.9563 | 0.9570 | 0.9566 |
|
| 86 |
+
| 0.1461 | 14.8186 | 2400 | 0.1171 | 0.9588 | 0.9588 | 0.9588 | 0.9588 |
|
| 87 |
+
| 0.133 | 15.4341 | 2500 | 0.1165 | 0.9548 | 0.9546 | 0.9549 | 0.9548 |
|
| 88 |
+
| 0.1258 | 16.0496 | 2600 | 0.1302 | 0.9480 | 0.9474 | 0.9500 | 0.9480 |
|
| 89 |
+
| 0.115 | 16.6698 | 2700 | 0.1320 | 0.9534 | 0.9537 | 0.9552 | 0.9534 |
|
| 90 |
+
| 0.1134 | 17.2853 | 2800 | 0.1171 | 0.9552 | 0.9549 | 0.9562 | 0.9552 |
|
| 91 |
+
| 0.1069 | 17.9054 | 2900 | 0.1127 | 0.9597 | 0.9595 | 0.9602 | 0.9597 |
|
| 92 |
+
|
| 93 |
+
|
| 94 |
+
### Framework versions
|
| 95 |
+
|
| 96 |
+
- Transformers 4.48.3
|
| 97 |
+
- Pytorch 2.6.0+cu124
|
| 98 |
+
- Datasets 3.4.0
|
| 99 |
+
- Tokenizers 0.21.0
|