NotFoundError: Key D/d1/d0/biases not found in checkpoint

[BernardoGO]

I'm trying to run the test with the pretrained checkpoint.

Unzipped pretrained.zip to GazeCorrection/log3_25_1/checkpoints.
When I run bash scripts/test_log20_3_25_1.sh this error is thrown:

NotFoundError (see above for traceback): Key D/d1/d0/biases not found in checkpoint
	 [[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]

[AliRashidnejad]

Hey @BernardoGO , this is happenning because of two below lines where the code tries to find latest checkpoints and start training on top of them, I guess pretrained which authores released is kind of buggy/incomplete and hence you need to retrain the model from scratch:
so comment lines 160-161 of GazeGAN.py and proceed training:

if ckpt and ckpt.model_checkpoint_path: pass # start_step = int(ckpt.model_checkpoint_path.split('model_', 2)[1].split('.', 2)[0]) # self.saver.restore(sess, ckpt.model_checkpoint_path) else: print("")