hezarai
/

crnn-base-fa-v2

Model card Files Files and versions Community

arxyzan commited on Apr 14

Commit

543881f

·

verified ·

1 Parent(s): 8697699

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -10,7 +10,7 @@ pipeline_tag: image-to-text
 ---
 A CRNN model for Persian OCR. This model is based on a simple CNN + LSTM architecture inspired by [this paper](https://arxiv.org/abs/1507.05717).
-This is a successor model to our previous model [hezarai/crnn-base-fa-64x256](https://huggingface.co/hezarai/crnn-base-fa-64x256).
 The improvements include:
 - 5X larger dataset
 - Change input image size from 64x256 to 32x384
@@ -29,7 +29,7 @@ pip install hezar
 ```python
 from hezar.models import Model
-crnn = Model.load("hezarai/crnn-fa-printed-96-long")
 texts = crnn.predict(["sample_image.jpg"])
 print(texts)
 ```

 ---
 A CRNN model for Persian OCR. This model is based on a simple CNN + LSTM architecture inspired by [this paper](https://arxiv.org/abs/1507.05717).
+This is a successor model to our previous model [hezarai/crnn-base-fa-v1](https://huggingface.co/hezarai/crnn-base-fa-v1).
 The improvements include:
 - 5X larger dataset
 - Change input image size from 64x256 to 32x384
 ```python
 from hezar.models import Model
+crnn = Model.load("hezarai/crnn-base-fa-v2")
 texts = crnn.predict(["sample_image.jpg"])
 print(texts)
 ```