Greg

Add support for loading models from google drive in the `ModelRegistry` Possible syntax. ``` ModelRegistry.register_provider("gdrive") _name_or_path = "rms/layoutlmv3-large-corr-ner" kwargs = { "provider":"gdrive", "__model_path__": __model_path__ } _name_or_path = ModelRegistry.get_local_path(_name_or_path, **kwargs) ```

Currently basic load balancing is used for work distributions, this needs to be replaces with much more robust solution. In original jina-ai work distribution is done via gRpc, this does...

Need to add webhook integration. Endpoints : - Add a webhook - Get all webhooks - Update webhook - Delete webhook - Get all logs for requests sent out for...

Implement health check registry. Registry needs to support following types: - http - script - tcp - sql Example configuration: **monitor.json** ``` { "shell" :"/bin/bash", "args": [ "./monitor/memory.sh" ], "id":...

To improve model performance during CPU inference we can convert the models for ONNX and then use if onnxruntime is available during inference time. Following script `check_onnx_runtime.py` can be used...

Implement TextFuseNet text box detection [TextFuseNet: Scene Text Detection with Richer Fused Features](https://www.ijcai.org/proceedings/2020/72)](https://github.com/ying09/TextFuseNet)

Text rendering that allows us to keep layout for generate text document.

There is a need for a job monitoring console that allows us to see what is going on in the system.

There is a need for webhook integration into the ICR server to allow for non message-driven communication. ``` Client ------> Async Processing Client

Status endpoint The server can return the following HTTP status codes: **200** – successful method call **4xx** – incorrect parameters of the method **5xx** – an error on the server-side...