wvsharber
commented
4 years ago Hi!
This is a known error that pops up occasionally. I made a hot fix for it, but I haven't pushed that change to GitHub yet. The easiest thing to do is try running it again, that usually fixes the problem. If you're finding that it keeps happening, you might try changing the value for K.
Basically what is happening is the song segmentation is looking to use the last five "moves" to produce a new sequence for the next song segment. If the model hasn't seen the previous five moves before, then it doesn't know what to do and gives this error. Since there is some randomness to the mapping, usually rerunning will produce a new map that doesn't cause the error.
Hope that helps and I'll try to push those changes soon.
nbanyan
I don't know if this is related to the FileNotFoundError I get when using other models.
(base) D:\Downloads\Utilities\BeatMapSynthesizer-master>python beatmapsynth.py "Music/Ehrling - Sthlm-Sunset.wav" "Sthlm Sunset by Ehrling" Hard segmented_HMM E:\anaconda3\lib\site-packages\pydub\utils.py:170: RuntimeWarning: Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work warn("Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work", RuntimeWarning) Loading Song... Song loaded successfully! Mapping with segmented Hidden Markov Model... Traceback (most recent call last): File "beatmapsynth.py", line 719, in
beat_map_synthesizer(args.song_path, args.song_name, args.difficulty, args.model, args.k, args.version)
File "beatmapsynth.py", line 48, in beat_map_synthesizer
segmented_HMM_mapper(song_path, song_name, difficulty, k = k, version = version)
File "beatmapsynth.py", line 330, in segmented_HMM_mapper
notes_list = segmented_HMM_notes_writer(y, sr, k, difficulty, version)
File "beatmapsynth.py", line 468, in segmented_HMM_notes_writer
preds = segment_predictions(segments_df, MC)
File "beatmapsynth.py", line 434, in segment_predictions
pred = HMM_model.walk(init_state = tuple(preds.iloc[-5:, 0]))
File "E:\anaconda3\lib\site-packages\markovify\chain.py", line 137, in walk
return list(self.gen(init_state))
File "E:\anaconda3\lib\site-packages\markovify\chain.py", line 126, in gen
next_word = self.move(state)
File "E:\anaconda3\lib\site-packages\markovify\chain.py", line 112, in move
choices, weights = zip(*self.model[state].items())
KeyError: ('0,2,0,7,1,3,0,7,999,999,999,999', '0,0,0,6,999,999,999,999,999,999,999,999', '0,1,2,5,1,2,0,6,999,999,999,999', '0,2,0,7,999,999,999,999,999,999,999,999', '999,999,999,999,1,3,1,3,999,999,999,999')