Closed rad182 closed 7 years ago
Hi @rad182, I'm also trying to do a simple example training using unseen_speakers. It sounds like I'm taking the same steps as you, though I'm using 17 training videos and 17 validation videos. I also encountered this error and will post here if I figure anything out.
I see you changed the title of the issue. Does that mean you figured out the problem? If so, could you post how you resolved it?
@michiyosony i just changed the minibatch_size
param in train() to the number of videos to train and it should work but I'm getting a new error which is updated above.
@rad182 I just saw that, thanks! For the benefit of future users, I think it would be good to revert this above description and open a new issue with the new stack track.
Please tell me if the the following is the correct way of making changes in train.py, to include vface and face_predictor_path ?
lip_gen = BasicGenerator(dataset_path=DATASET_DIR, vtype = "face", face_predictor_path="/home/deepakgupta1313/Desktop/Deepak/Programs/Github/LipNet/common/predictors/shape_predictor_68_face_landmarks.dat", minibatch_size=minibatch_size, img_c=img_c, img_w=img_w, img_h=img_h, frames_n=frames_n,
FYI if the validation_steps == 0 will raise this ValueError
I have edited fit_generator and assign a value to validation_step, is this the correct way to so it @rizkiarm @shensheng27
lipnet.model.fit_generator(generator=lip_gen.next_train(),
steps_per_epoch=lip_gen.default_training_steps, epochs=stop_epoch,
validation_data=lip_gen.next_val(), validation_steps=1,
callbacks=[checkpoint, statistics, visualize, lip_gen, tensorboard, csv_logger],
initial_epoch=start_epoch,
verbose=1,
max_q_size=5,
workers=2,
pickle_safe=True)
Hi, Thanks for this open sourcing this lipnet. I'm still new to ML and still learning. However I'm trying to train 10 videos (from grid) for now(just for testing) using the "unseen_speakers" and getting an error about validation steps. Also enabled vtype='face', face_predictor_path=FACE_PREDICTOR_PATH
Below is the complete log