int8 ERROR when I run convert_tflite.py

[aluds123]

I run this command:
python convert_tflite.py --weights ./checkpoints/yolov4-416 --output ./checkpoints/yolov4-416-int8.tflite --quantize_mode int8 --dataset ./coco_dataset/coco/val207.txt

And occur:
calibration image /home/soc507/darknet/data/coco/images/val2014//COCO_val2014_000000183693.jpg
calibration image /home/soc507/darknet/data/coco/images/val2014//COCO_val2014_000000150834.jpg
calibration image /home/soc507/darknet/data/coco/images/val2014//COCO_val2014_000000140270.jpg
I0907 13:29:51.712005 139893978974016 convert_tflite.py:48] model saved to: ./checkpoints/yolov3-tiny-int8.tflite
Traceback (most recent call last):
File "convert_tflite.py", line 76, in
app.run(main)
File "/home/aluds/anaconda3/envs/tf230/lib/python3.5/site-packages/absl/app.py", line 300, in run
_run_main(main, args)
File "/home/aluds/anaconda3/envs/tf230/lib/python3.5/site-packages/absl/app.py", line 251, in _run_main
sys.exit(main(argv))
File "convert_tflite.py", line 72, in main
demo()
File "convert_tflite.py", line 52, in demo
interpreter.allocate_tensors()
File "/home/aluds/anaconda3/envs/tf230/lib/python3.5/site-packages/tensorflow/lite/python/interpreter.py", line 243, in allocate_tensors
return self._interpreter.AllocateTensors()
RuntimeError: tensorflow/lite/kernels/dequantize.cc:61 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 64 (DEQUANTIZE) failed to prepare.

Question:
Can someone give me some suggestion? Thank you.

[BernardinD]

Answers to this issue? I'm having the same one

[drahmad89]

Answers to this issue? I'm having the same one

[4yougames]

Can someone give me some suggestion? Thank you.

[YLTsai0609]

A similar issue occurred to me, I use yolov3. Any suggestions?

[istomoya]

Fix in
#214
worked in converting yolov3-tiny to TF-Lite int8.
I had to comment out lines in save_model.py and add the lines to detect.py.