转换通过darknet框架训练自有数据集后的.weights文件，出现错误

Eve66666 commented 4 years ago

你好，通过指令：python convert_tflite.py --weights ./data/yolov4.weights --output ./data/yolov4-fp16.tflite --quantize_mode float16 我想把一个在darknet框架下训练得到的.weights权重文件转为tflite文件，但是出现了如下问题： ValueError: cannot reshape array of size 4559937 into shape (1024,512,3,3) 具体报错内容如下

Traceback (most recent call last):
  File "convert_tflite.py", line 109, in <module>
    app.run(main)
  File "E:\anaconda\envs\python-train\lib\site-packages\absl\app.py", line 299, in run
    _run_main(main, args)
  File "E:\anaconda\envs\python-train\lib\site-packages\absl\app.py", line 250, in _run_main
    sys.exit(main(argv))
  File "convert_tflite.py", line 104, in main
    save_tflite()
  File "convert_tflite.py", line 59, in save_tflite
    utils.load_weights(model, FLAGS.weights)
  File "E:\11yolov4\tensorflow-yolov4-tflite\tensorflow-yolov4-tflite-master\core\utils.py", line 114, in load_weights
    conv_weights = conv_weights.reshape(conv_shape).transpose([2, 3, 1, 0])
ValueError: cannot reshape array of size 4559937 into shape (1024,512,3,3)

请问如何可以解决这个问题？