int8 quantization: op_context.input->type == kTfLiteUInt8 || op_context.input->type == kTfLiteInt8 || op_context.input->type == kTfLiteInt16 || op_context.input->type == kTfLiteFloat16 was not true.Node number 15 (DEQUANTIZE) failed to prepare

Colab에서 Yolov4를 tensorlite int8로 변환하는데, 위와 같은 에러가 뜨네요.....

int8 quantization: op_context.input->type == kTfLiteUInt8 || op_context.input-type == kTfLiteInt8 || op_context.input->type == kTfLiteInt16 || op_context.input->type == kTfLiteFloat16 was not true.Node number 15 (DEQUANTIZE) failed to prepare

그래서 열심히 뒤져보니, 해결책이 아래에 있었습니다.

FYI, I ran your colab with !pip install tf-nightly transformers==4.2.2. It worked without any errors.

개발자분들 화이팅입니다.