Hugging Face's logo

yumjunstar
/
trocr-small-printedkorean-deleteunusedchar

Image-Text-to-Text
Transformers
PyTorch
vision-encoder-decoder
Generated from Trainer
Inference Endpoints
Model card Files Community
1
yumjunstar commited on
Commit
b39c871
·
1 Parent(s): 0a226a0

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +85 -0
README.md ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - wer
7
+ - accuracy
8
+ model-index:
9
+ - name: trocr-small-printedkorean-deleteunusedchar
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # trocr-small-printedkorean-deleteunusedchar
17
+
18
+ This model is a fine-tuned version of [team-lucid/trocr-small-korean](https://huggingface.co/team-lucid/trocr-small-korean) on an unknown dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 2.1253
21
+ - Cer: 0.3884
22
+ - Wer: 0.4405
23
+ - Accuracy: 27.0903
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 4e-05
43
+ - train_batch_size: 128
44
+ - eval_batch_size: 192
45
+ - seed: 42
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: linear
48
+ - num_epochs: 10
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Cer | Wer | Accuracy |
54
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:--------:|
55
+ | 3.5404 | 0.43 | 1000 | 3.5688 | 0.5737 | 0.6446 | 17.7258 |
56
+ | 3.365 | 0.85 | 2000 | 3.4212 | 0.5664 | 0.6546 | 18.0602 |
57
+ | 3.266 | 1.28 | 3000 | 3.3538 | 0.5637 | 0.6815 | 18.7291 |
58
+ | 3.2922 | 1.71 | 4000 | 3.3051 | 0.5613 | 0.7161 | 18.0602 |
59
+ | 3.1287 | 2.13 | 5000 | 3.2649 | 0.5577 | 0.7164 | 18.7291 |
60
+ | 3.0906 | 2.56 | 6000 | 3.2016 | 0.5572 | 0.6139 | 18.3946 |
61
+ | 3.0015 | 2.99 | 7000 | 3.0851 | 0.5530 | 0.6018 | 18.3946 |
62
+ | 2.7418 | 3.42 | 8000 | 2.8927 | 0.5264 | 0.6052 | 19.3980 |
63
+ | 2.5107 | 3.84 | 9000 | 2.6466 | 0.4821 | 0.5881 | 19.3980 |
64
+ | 2.3091 | 4.27 | 10000 | 2.4865 | 0.4545 | 0.5651 | 20.4013 |
65
+ | 2.1395 | 4.7 | 11000 | 2.3749 | 0.4336 | 0.5323 | 21.7391 |
66
+ | 2.0453 | 5.12 | 12000 | 2.3070 | 0.4222 | 0.5157 | 21.7391 |
67
+ | 2.0421 | 5.55 | 13000 | 2.2667 | 0.4143 | 0.5001 | 24.0803 |
68
+ | 2.0085 | 5.98 | 14000 | 2.2201 | 0.4060 | 0.4809 | 24.0803 |
69
+ | 1.8766 | 6.4 | 15000 | 2.1988 | 0.4026 | 0.4766 | 23.7458 |
70
+ | 1.8987 | 6.83 | 16000 | 2.1788 | 0.3969 | 0.4658 | 24.4147 |
71
+ | 1.8921 | 7.26 | 17000 | 2.1702 | 0.3955 | 0.4581 | 24.4147 |
72
+ | 1.8253 | 7.69 | 18000 | 2.1554 | 0.3938 | 0.4513 | 26.7559 |
73
+ | 1.8281 | 8.11 | 19000 | 2.1481 | 0.3925 | 0.4534 | 26.0870 |
74
+ | 1.8751 | 8.54 | 20000 | 2.1419 | 0.3905 | 0.4500 | 26.0870 |
75
+ | 1.8226 | 8.97 | 21000 | 2.1317 | 0.3905 | 0.4439 | 26.7559 |
76
+ | 1.813 | 9.39 | 22000 | 2.1285 | 0.3895 | 0.4407 | 26.0870 |
77
+ | 1.7472 | 9.82 | 23000 | 2.1253 | 0.3884 | 0.4405 | 27.0903 |
78
+
79
+
80
+ ### Framework versions
81
+
82
+ - Transformers 4.28.0
83
+ - Pytorch 1.13.1+cu116
84
+ - Datasets 2.14.4
85
+ - Tokenizers 0.13.3