Update README.md
Browse files
README.md
CHANGED
|
@@ -4,36 +4,18 @@ model-index:
|
|
| 4 |
- name: sinhala-ocr-model-v2
|
| 5 |
results: []
|
| 6 |
pipeline_tag: image-to-text
|
| 7 |
-
widget:
|
| 8 |
-
- src: >-
|
| 9 |
-
https://datasets-server.huggingface.co/assets/Ransaka/sinhala_synthetic_ocr/--/bf7c8a455b564cd73fe035031e19a5f39babb73b/--/default/train/0/image/image.jpg
|
| 10 |
-
example_title: Synthetic 1
|
| 11 |
-
- src: >-
|
| 12 |
-
https://datasets-server.huggingface.co/assets/Ransaka/sinhala_synthetic_ocr/--/bf7c8a455b564cd73fe035031e19a5f39babb73b/--/default/train/1/image/image.jpg
|
| 13 |
-
example_title: Synthetic 2
|
| 14 |
-
- src: >-
|
| 15 |
-
https://datasets-server.huggingface.co/assets/Ransaka/sinhala_synthetic_ocr/--/bf7c8a455b564cd73fe035031e19a5f39babb73b/--/default/train/9/image/image.jpg
|
| 16 |
-
example_title: Synthetic 3
|
| 17 |
---
|
| 18 |
|
| 19 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 20 |
should probably proofread and complete it, then remove this comment. -->
|
| 21 |
|
| 22 |
-
#
|
| 23 |
|
| 24 |
-
|
| 25 |
-
It achieves the following results on the evaluation set:
|
| 26 |
-
- eval_loss: 4.8494
|
| 27 |
-
- eval_cer: 0.4227
|
| 28 |
-
- eval_runtime: 229.7041
|
| 29 |
-
- eval_samples_per_second: 1.776
|
| 30 |
-
- eval_steps_per_second: 0.444
|
| 31 |
-
- epoch: 5.23
|
| 32 |
-
- step: 400
|
| 33 |
|
| 34 |
## Model description
|
| 35 |
|
| 36 |
-
|
| 37 |
|
| 38 |
## Intended uses & limitations
|
| 39 |
|
|
@@ -43,25 +25,12 @@ More information needed
|
|
| 43 |
|
| 44 |
More information needed
|
| 45 |
|
| 46 |
-
## Training procedure
|
| 47 |
-
|
| 48 |
-
### Training hyperparameters
|
| 49 |
-
|
| 50 |
-
The following hyperparameters were used during training:
|
| 51 |
-
- learning_rate: 1e-05
|
| 52 |
-
- train_batch_size: 4
|
| 53 |
-
- eval_batch_size: 4
|
| 54 |
-
- seed: 42
|
| 55 |
-
- gradient_accumulation_steps: 4
|
| 56 |
-
- total_train_batch_size: 16
|
| 57 |
-
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 58 |
-
- lr_scheduler_type: linear
|
| 59 |
-
- training_steps: 6000
|
| 60 |
-
- mixed_precision_training: Native AMP
|
| 61 |
-
|
| 62 |
### Framework versions
|
| 63 |
|
| 64 |
- Transformers 4.35.2
|
| 65 |
- Pytorch 2.0.0
|
| 66 |
- Datasets 2.16.0
|
| 67 |
-
- Tokenizers 0.15.0
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
- name: sinhala-ocr-model-v2
|
| 5 |
results: []
|
| 6 |
pipeline_tag: image-to-text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
---
|
| 8 |
|
| 9 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 10 |
should probably proofread and complete it, then remove this comment. -->
|
| 11 |
|
| 12 |
+
# TrOCR-Sinhala
|
| 13 |
|
| 14 |
+
See training metrics tab for performance details.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
## Model description
|
| 17 |
|
| 18 |
+
This model is finetuned version of Microsoft [TrOCR Printed](https://huggingface.co/microsoft/trocr-base-printed)
|
| 19 |
|
| 20 |
## Intended uses & limitations
|
| 21 |
|
|
|
|
| 25 |
|
| 26 |
More information needed
|
| 27 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
### Framework versions
|
| 29 |
|
| 30 |
- Transformers 4.35.2
|
| 31 |
- Pytorch 2.0.0
|
| 32 |
- Datasets 2.16.0
|
| 33 |
+
- Tokenizers 0.15.0
|
| 34 |
+
|
| 35 |
+
## Examples
|
| 36 |
+
<img src='https://datasets-server.huggingface.co/assets/Ransaka/sinhala_synthetic_ocr/--/bf7c8a455b564cd73fe035031e19a5f39babb73b/--/default/train/0/image/image.jpg'>
|