hz

> Any luck solving this?

> I have the same issue, I am consistently getting 100% accuracy on the validation dataset, has anyone been able to solve this? Any luck solving this?

thanks ---Original--- From: ***@***.***> Date: Wed, Jun 22, 2022 20:10 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [4uiiurz1/keras-arcface] why I get a 100% acc on the valid data?(#14) No, no...

> No, no luck in solving it, but I came across the following article [https://yiwenlai.medium.com/how-to-train-with-arcface-loss-to-improve-model-classification-accuracy-d4035195aeb9](url). It mentions about the 100% accuracy error and how to solve it. Please go through...

2 ways ，it works    1.use tf 2+  2. trans speech brain ArcFace to tf version ---Original--- From: "Kevin ***@***.***> Date: Sun, Aug 28, 2022 06:55 AM To: ***@***.***>; Cc:...

> any numbers on `much worse` i.e., wer comparison between streaming and non-streaming ? some test wavs,no-streaming decoding sentence acc is 0.96 and wer is 0.9. but streaming-decoding is 0.78...

sanic: websocket send a json message C language: recv messga,got some more info eg: send {"name":"sanic"} and recv :{"name":"sanic"}kkfsdfs} now for sanic, send json + "\0" works