update

README.md

@@ -66,10 +66,10 @@ Then go to `http://localhost:8501` in your browser to run TexTeller in the web.
 We use [ray serve](https://github.com/ray-project/ray) to provide a simple API for using TexTeller in your own projects. To start the server, navigate to the `TexTeller/src` directory and run the following command:
 
 ```bash
-python serve.py  # default settings
+python server.py  # default settings
 ```
 
-You can pass the following arguments to the `serve.py` script to get custom inference settings(e.g. `python serve.py --use_gpu` to enable GPU inference):
+You can pass the following arguments to the `server.py` script to get custom inference settings(e.g. `python server.py --use_gpu` to enable GPU inference):
 
 | Argument | Description |
 | --- | --- |

requirements.txt

@@ -7,3 +7,4 @@ ray[serve]
 accelerate
 tensorboardX
 nltk
+python-multipart

src/client_demo.py

@@ -7,8 +7,4 @@ with open(img_path, 'rb') as img:
     files = {'img': img}
     response = requests.post(url, files=files)
 
-# data = {"img_path": img_path}
-
-# response = requests.post(url, json=data)
-
 print(response.text)

src/models/ocr_model/utils/inference.py

@@ -21,6 +21,7 @@ def inference(
     if isinstance(imgs_path[0], str):
         imgs = convert2rgb(imgs_path) 
     else:  # already numpy array(rgb format)
+        assert isinstance(imgs_path[0], np.ndarray)
         imgs = imgs_path 
     imgs = inference_transform(imgs)
     pixel_values = torch.stack(imgs)

src/server.py

@@ -67,6 +67,7 @@ class Ingress:
         img_rb = await form['img'].read()
 
         img_nparray = np.frombuffer(img_rb, np.uint8)
+        img_nparray = cv2.imdecode(img_nparray, cv2.IMREAD_COLOR)
         img_nparray = cv2.cvtColor(img_nparray, cv2.COLOR_BGR2RGB)
         pred = await self.texteller_server.predict.remote(img_nparray)
         return pred