10 ms·

Show HN: DeepRhyme (D-Prime) – Generating dope rhymes with machine learning

208 points by sweezyjeezy 10y ago

ada1981 10y ago
I'd like to see you team up with some AI Beat generators and then use text-to speech and produce albums.
random_gangster 10y ago
Stupidest thing I've ever seen. You think you can just press and measure some correlations and give something a cool name? Get out of DL bro.
jgotti92 10y ago
Awesome!
mobiuscog 10y ago
How will we know the AI versions ?
sjfiekzjndn 10y ago
I'm sensing an impending gangsta job shortage. I wonder if machines become better at being outlawz, then might these individuals pursue some form of higher street education to keep up?
jeyoor 10y ago
I wonder if approaches like this will work for poetry (or prose) from different cultures. I'd be especially interested to see if some of these techniques are effective on languages markedly different from English. (Maybe Chinese and Arabic?)
- DrNuke 10y ago
  For that, they should consider the underlying universal metrics (or meters) instead of the local languages, along the path derived by Prof Nigel Fabb in the UK https://www.amazon.co.uk/Meter-Poetry-Theory-Nigel-Fabb/dp/0521713250 https://www.amazon.co.uk/Meter-Poetry-Theory-Nigel-Fabb/dp/0... of which you have a review here https://web.stanford.edu/~kiparsky/Papers/fabb_halle_review.pdf https://web.stanford.edu/~kiparsky/Papers/fabb_halle_review....
Karunamon 10y ago
I have to giggle a little bit at the implications of taking source material that's just about every negative *ist word it's possible to be and then using ML to generate a version without that content.
- farright 10y ago
  I think this is where were supposed to invoke the intersectionality fairy.
0xdeadbeefbabe 10y ago
Somehow these AI projects always seem to demonstrate how real intelligence is better.
- singham 10y ago
  Baby steps. We all were babies once.
singham 10y ago
Next stop : Pun recognizer or pun generator.
simplemath 10y ago
D-Prime is an incredibly good AI MC name Whoever thought that one up deserves free lunch in perpetuity
- mrdmnd 10y ago
  No Free Lunch ;)
jcoffland 10y ago
To the uninitiated these may seem like reasonable results because to the uninitiated hip hop is just a bunch of unintelligible words strung together anyway but these lyrics don't make much sense. There's no cohesive stream of thought behind the lines. It's a semirandom juxtaposition of phrases and a great example of one of the biggest limitations of deep learning softwares. Even the best are only able to maintain very limited context. This is not just a matter of scaling up to larger networks. Current methods require exponentially larger networks to achieve linear increases in context handling. So then why can our brains do it? The answer is simply that our brains are more than neural networks.
- sweezyjeezy 10y ago
  Author here. You would be surprised how much of the input lyrics were just semirandom juxtaposition of phrases. The best hip hop is not this, sure, but some of it certainly is.
  - croon 10y ago
    Sure, and most human made music is crap. That's not a diss or putting down your work, which is laudable, but the distinction GP presented I believe stands. It's very hard to produce _great_ content this way, simply because there are a lot more variables and dimensions to writing good music and lyrics.
- erikschoster 10y ago
  Rap music says something. It's hard not to read this project as just a further disenfranchisement its authors.
  - sweezyjeezy 10y ago
    Author here. I'm sad that you think that. I love hip hop, and I know how far away this model is from the best MCs. The model has no idea what words actually mean, it's just spitting out a grammar it has learned. We are nowhere near the point where an AI can formulate an original thought and articulate it in rhyme, and I even called my model a 'parlour trick' in the post. It was just supposed to be fun.
uniclaude 10y ago
Don't know how good you are with cache invalidation But your naming game is tight, that's without a question
- corey_moncure 10y ago
  I think you're on to something here! It might be a while before we train a NN that can out-rap the pros. But can we raise the bar for applying the style to alternative content domains to generate educational raps that aren't lame?
aantix 10y ago
Looking forward to the time When Siri can spit Some dope rhymes
to-mi 10y ago
A rap lyrics generation algorithm with deep learning was also recently published [1] in KDD conference and it also got quite a lot of publicity in other media: see the associated website at http://deepbeat.org http://deepbeat.org. [1] Malmi et al.: DopeLearning: A Computational Approach to Rap Lyrics Generation http://www.kdd.org/kdd2016/papers/files/adf0399-malmiA.pdf http://www.kdd.org/kdd2016/papers/files/adf0399-malmiA.pdf
- sweezyjeezy 10y ago
  Author of DeepRhyme here. With respect to the authors of DeepBeat, what they are doing is less ambitious. They are taking full existing lines out of a rap lyric corpus and assembling it into a verse that rhymes and makes sense. There is a paper that is very similar to what I did : http://www.emnlp2015.org/proceedings/EMNLP/pdf/EMNLP221.pdf http://www.emnlp2015.org/proceedings/EMNLP/pdf/EMNLP221.pdf . It's hard to compare our models, because they don't give much output text. They trained on 1% of the data I did, so I'm a bit dubious how successful they could have been. I am writing a followup post where I'm going to talk about previous work, I hope no one takes any disrespect.
  - tigeba 10y ago
    Instead of completely banning actual lines from other material, it might be interesting to allow D-Prime to quote or slightly modify a phrase if it met some high threshold of notability.
  - YeGoblynQueenne 10y ago
    >> It's hard to compare our models, because they don't give much output text. The referenced paper doesn't give many examples of output, but the authors have made the system available online on the url given above.
    - sweezyjeezy 10y ago
      I think you're talking about DeepBeat, they aren't tackling the same problem, as I mentioned.
david-given 10y ago
I really want to drop in a Shakespeare play or two into the dataset and see what comes out. MC MacB comin for yo, Dun-king.
- FroshKiller 10y ago
  Mad Magazine (or maybe it was Cracked) did a pretty good rap parody of Macbeth years ago, maybe the late '80s or early '90s.
gwbas1c 10y ago
I seriously thought I was going to click on that link and see the stereotypical cliche rap: Yo yo yo I'm Deep Rhyme and I'm hear to say I figured out how to rap in a whole new way! My name's D Prime and I'm an A-I Maybe some day I'll calculate all of Pi! Deep Rhyme in the house, yo! (mike drop)
- bbctol 10y ago
  To be fair, that would be a far more impressive result than this!
david-given 10y ago
...does enough chap-hop exist to form a decent dataset? https://www.youtube.com/watch?v=0iRTB-FTMdk https://www.youtube.com/watch?v=0iRTB-FTMdk
michaelbuckbee 10y ago
Linked (but a little buried) in the post is this absolute standout of a YouTube video about the rhyme scheme in Eminem's Lose Yourself. https://www.youtube.com/watch?v=ooOL4T-BAg0 https://www.youtube.com/watch?v=ooOL4T-BAg0 Really drives home both how much more sophisticated hip hop has become and the incredible degree of difficulty in trying to simulate such sophisticated language structures.
- dancek 10y ago
  Eminem has similar trickery going on in most songs and that's probably one of the reasons of his success. Of course many other rappers do the same, and surely Eminem isn't the inventor of funky rhyme patterns. It's been 15 years but I used to analyze the patterns. Brings back memories.
YeGoblynQueenne 10y ago
For a moment there I thought that thing had learned to rhyme on its own. Then I figured out it's got some rules that help it along. Phew. That'd have been scary.
daveytea 10y ago
This is great. Well done to OP! I've been considering creating something like this in the future as well. You mentioned that you used the lyrics from 50,000 rap songs. Did you have any filters for quality or certain sub-genres? What era were the rap songs from? Maybe the 'quality' would be higher if the data set was narrowed to what you consider quality hip hop lyrics. Also from your post: "You can see that D-Prime does internally rhyme to a small extent, but making the model better in this regard seemed like it was going to be tough, and I didn’t pursue it." What were your main challenges to improving/pursing better internal rhyme schemes?
- FilterSweep 10y ago
  > Maybe the 'quality' would be higher if the data set was narrowed to what you consider quality hip hop lyrics. Also consider the slang in different eras has changed over time, too. The example I often use is the phrase "word is bond" in the 90s which was replaced by "on God" in the 00s. I wonder if you can get D-Prime trained off separate datasets.
- sweezyjeezy 10y ago
  Hey man, thanks! I didn't try to filter to any subgenres, but I did try to give the model as easy a time as possible by trying to stick verses which stuck to the most common words (lowering the perplexity of the corpus). I would call this genre 'mediocre', and may explain why I had such a hard time with the sexism thing. Internal rhyming / assonance and other stuff could be done by rewarding the model for using it during the beam search. I suspect it's going to be hard to tune how to set the reward so it still makes sense, it only just makes sense already. I also think at this point you would really need to sit down and figure out what rhymes and what doesn't. I did the quickest, hackiest thing I could think of, which was good enough for generating end rhymes, but if you want to have internal rhymes and not massively constrain the model, I think you need to work it out purely from analysing the syllables.
kastnerkyle 10y ago
OP and others may be interested in this approach by Sony CSL from ~2012 using constraint Markov chains for matching lyrical style with explicit rhyme constraints, including (e.g.) Bob Dylan style with constraint of Beatles' Yesterday [0]. This is the precursor to the work that brought on "Daddy's Car"[1], though the techniques appear to have changed a bit reading their recent publications. I find their paper on recent approaches for music synthesis with style matching or other constraints [2] quite readable, and really good! I am hoping to build out a version of what is described by the paper. There is also a very nice talk by F. Pachet on this subject here [3], and his older videos are good too (they should be in the recommendations of YouTube). Still reading about DeepRhyme, but it seems great. Also a fantastic writeup! [0] https://www.csl.sony.fr/downloads/papers/2012/barbieri-12a.pdf https://www.csl.sony.fr/downloads/papers/2012/barbieri-12a.p... [1] https://www.youtube.com/watch?v=LSHZ_b05W7o https://www.youtube.com/watch?v=LSHZ_b05W7o [2] https://arxiv.org/abs/1609.05152 https://arxiv.org/abs/1609.05152 [3] https://www.youtube.com/watch?v=5K4hn6cBUPU https://www.youtube.com/watch?v=5K4hn6cBUPU
- sweezyjeezy 10y ago
  Hey thanks! I will check this out tomorrow!
adamsea 10y ago
As a huge hip-hop head, I will just add that there are two aspects to evaluating rhymes: How it sounds when delivered by the artist, and how it reads. The best rap succeeds in both domains. Bob Dylan won the Nobel Prize because he succeeds in both domains. In this case, we obviously can only consider the second domain, how it reads. So we should evaluate the results exactly how we would evaluate any other poem (while giving due respect to the conventions and tropes of the form, ie, rap has its own classic metaphors and themes).
infinitone 10y ago
This is awesome! I've been meaning to do something like this for awhile.. great that someone else finally did it. Any chance you'd open source it?
casoninabox 10y ago
First, this is awesome. Great results. I have been working on exactly this for the past few weeks. I went ahead and produced a song and a lyric video from mine (I call it Lil' Markov). It started out because I noticed patterns in Lil Wayne song where he'd say some idiom and then rhyme it. It seemed so predictable I figured I could make a bot that automated Lil Wayne's process... From what I can tell our methods are very similar. I have a thing that allows you to input a concept and it will try to rap around that concept and a `whiteGuilt()` function (that's whats up with all the ninjas), but other than that our process is 90% similar. We should collaborate on something! Message me. Version 0: https://soundcloud.com/cason-clagg/aesopxcasonxmarkov https://soundcloud.com/cason-clagg/aesopxcasonxmarkov Version 1: https://www.youtube.com/watch?v=fmcrjH6BKag https://www.youtube.com/watch?v=fmcrjH6BKag Version 2: (No music yet, bit here's some sample lyrics based on the concept "beef") FIND MYSELF WITH MILLIONS THE PIGGIES THAT'LL SHARE HOPPIN' AROUND SATURN'S RING I'M JUICED OFF DOG DARE A DESCRIPTIVE LYRICIST LOUNGING IN LOVE LIKE YOU AIN'T EAGER TO KILL EACH ONE OF THE DOVES HE RAN FOR A BEAST HE'S A PARAGON OF YORE DIDN'T THINK OF HIS BLOOD ON THE FACTORY FLOOR
lifeformed 10y ago
Next, you should plug this into a text-to-speech program and play it with a simple beat generator.
grandalf 10y ago
It would be fascinating to try training it on some HRC and Trump speeches. I am doubtful most people would be able to tell the difference between generated remarks and the real thing.

208 points by sweezyjeezy 10y ago

ada1981 10y ago

I'd like to see you team up with some AI Beat generators and then use text-to speech and produce albums.

Stupidest thing I've ever seen. You think you can just press and measure some correlations and give something a cool name? Get out of DL bro.

jgotti92 10y ago

Awesome!

mobiuscog 10y ago

How will we know the AI versions ?

sjfiekzjndn 10y ago

I'm sensing an impending gangsta job shortage. I wonder if machines become better at being outlawz, then might these individuals pursue some form of higher street education to keep up?

jeyoor 10y ago

I wonder if approaches like this will work for poetry (or prose) from different cultures. I'd be especially interested to see if some of these techniques are effective on languages markedly different from English. (Maybe Chinese and Arabic?)

DrNuke 10y ago
For that, they should consider the underlying universal metrics (or meters) instead of the local languages, along the path derived by Prof Nigel Fabb in the UK https://www.amazon.co.uk/Meter-Poetry-Theory-Nigel-Fabb/dp/0521713250 https://www.amazon.co.uk/Meter-Poetry-Theory-Nigel-Fabb/dp/0... of which you have a review here https://web.stanford.edu/~kiparsky/Papers/fabb_halle_review.pdf https://web.stanford.edu/~kiparsky/Papers/fabb_halle_review....

Karunamon 10y ago

I have to giggle a little bit at the implications of taking source material that's just about every negative *ist word it's possible to be and then using ML to generate a version without that content.

farright 10y ago
I think this is where were supposed to invoke the intersectionality fairy.

0xdeadbeefbabe 10y ago

Somehow these AI projects always seem to demonstrate how real intelligence is better.

singham 10y ago
Baby steps. We all were babies once.

singham 10y ago

Next stop : Pun recognizer or pun generator.

simplemath 10y ago

D-Prime is an incredibly good AI MC name Whoever thought that one up deserves free lunch in perpetuity

mrdmnd 10y ago
No Free Lunch ;)

jcoffland 10y ago

To the uninitiated these may seem like reasonable results because to the uninitiated hip hop is just a bunch of unintelligible words strung together anyway but these lyrics don't make much sense. There's no cohesive stream of thought behind the lines. It's a semirandom juxtaposition of phrases and a great example of one of the biggest limitations of deep learning softwares. Even the best are only able to maintain very limited context. This is not just a matter of scaling up to larger networks. Current methods require exponentially larger networks to achieve linear increases in context handling. So then why can our brains do it? The answer is simply that our brains are more than neural networks.

sweezyjeezy 10y ago
Author here. You would be surprised how much of the input lyrics were just semirandom juxtaposition of phrases. The best hip hop is not this, sure, but some of it certainly is.
- croon 10y ago
  Sure, and most human made music is crap. That's not a diss or putting down your work, which is laudable, but the distinction GP presented I believe stands. It's very hard to produce _great_ content this way, simply because there are a lot more variables and dimensions to writing good music and lyrics.
erikschoster 10y ago
Rap music says something. It's hard not to read this project as just a further disenfranchisement its authors.
- sweezyjeezy 10y ago
  Author here. I'm sad that you think that. I love hip hop, and I know how far away this model is from the best MCs. The model has no idea what words actually mean, it's just spitting out a grammar it has learned. We are nowhere near the point where an AI can formulate an original thought and articulate it in rhyme, and I even called my model a 'parlour trick' in the post. It was just supposed to be fun.

uniclaude 10y ago

Don't know how good you are with cache invalidation But your naming game is tight, that's without a question

corey_moncure 10y ago
I think you're on to something here! It might be a while before we train a NN that can out-rap the pros. But can we raise the bar for applying the style to alternative content domains to generate educational raps that aren't lame?

aantix 10y ago

Looking forward to the time When Siri can spit Some dope rhymes

to-mi 10y ago

A rap lyrics generation algorithm with deep learning was also recently published [1] in KDD conference and it also got quite a lot of publicity in other media: see the associated website at http://deepbeat.org http://deepbeat.org. [1] Malmi et al.: DopeLearning: A Computational Approach to Rap Lyrics Generation http://www.kdd.org/kdd2016/papers/files/adf0399-malmiA.pdf http://www.kdd.org/kdd2016/papers/files/adf0399-malmiA.pdf

sweezyjeezy 10y ago
Author of DeepRhyme here. With respect to the authors of DeepBeat, what they are doing is less ambitious. They are taking full existing lines out of a rap lyric corpus and assembling it into a verse that rhymes and makes sense. There is a paper that is very similar to what I did : http://www.emnlp2015.org/proceedings/EMNLP/pdf/EMNLP221.pdf http://www.emnlp2015.org/proceedings/EMNLP/pdf/EMNLP221.pdf . It's hard to compare our models, because they don't give much output text. They trained on 1% of the data I did, so I'm a bit dubious how successful they could have been. I am writing a followup post where I'm going to talk about previous work, I hope no one takes any disrespect.
- tigeba 10y ago
  Instead of completely banning actual lines from other material, it might be interesting to allow D-Prime to quote or slightly modify a phrase if it met some high threshold of notability.
- YeGoblynQueenne 10y ago
  >> It's hard to compare our models, because they don't give much output text. The referenced paper doesn't give many examples of output, but the authors have made the system available online on the url given above.
  - sweezyjeezy 10y ago
    I think you're talking about DeepBeat, they aren't tackling the same problem, as I mentioned.

david-given 10y ago

I really want to drop in a Shakespeare play or two into the dataset and see what comes out. MC MacB comin for yo, Dun-king.

FroshKiller 10y ago
Mad Magazine (or maybe it was Cracked) did a pretty good rap parody of Macbeth years ago, maybe the late '80s or early '90s.

gwbas1c 10y ago

I seriously thought I was going to click on that link and see the stereotypical cliche rap: Yo yo yo I'm Deep Rhyme and I'm hear to say I figured out how to rap in a whole new way! My name's D Prime and I'm an A-I Maybe some day I'll calculate all of Pi! Deep Rhyme in the house, yo! (mike drop)

bbctol 10y ago
To be fair, that would be a far more impressive result than this!

david-given 10y ago

...does enough chap-hop exist to form a decent dataset? https://www.youtube.com/watch?v=0iRTB-FTMdk https://www.youtube.com/watch?v=0iRTB-FTMdk

michaelbuckbee 10y ago

Linked (but a little buried) in the post is this absolute standout of a YouTube video about the rhyme scheme in Eminem's Lose Yourself. https://www.youtube.com/watch?v=ooOL4T-BAg0 https://www.youtube.com/watch?v=ooOL4T-BAg0 Really drives home both how much more sophisticated hip hop has become and the incredible degree of difficulty in trying to simulate such sophisticated language structures.

dancek 10y ago
Eminem has similar trickery going on in most songs and that's probably one of the reasons of his success. Of course many other rappers do the same, and surely Eminem isn't the inventor of funky rhyme patterns. It's been 15 years but I used to analyze the patterns. Brings back memories.

YeGoblynQueenne 10y ago

For a moment there I thought that thing had learned to rhyme on its own. Then I figured out it's got some rules that help it along. Phew. That'd have been scary.

daveytea 10y ago

This is great. Well done to OP! I've been considering creating something like this in the future as well. You mentioned that you used the lyrics from 50,000 rap songs. Did you have any filters for quality or certain sub-genres? What era were the rap songs from? Maybe the 'quality' would be higher if the data set was narrowed to what you consider quality hip hop lyrics. Also from your post: "You can see that D-Prime does internally rhyme to a small extent, but making the model better in this regard seemed like it was going to be tough, and I didn’t pursue it." What were your main challenges to improving/pursing better internal rhyme schemes?

FilterSweep 10y ago
> Maybe the 'quality' would be higher if the data set was narrowed to what you consider quality hip hop lyrics. Also consider the slang in different eras has changed over time, too. The example I often use is the phrase "word is bond" in the 90s which was replaced by "on God" in the 00s. I wonder if you can get D-Prime trained off separate datasets.
sweezyjeezy 10y ago
Hey man, thanks! I didn't try to filter to any subgenres, but I did try to give the model as easy a time as possible by trying to stick verses which stuck to the most common words (lowering the perplexity of the corpus). I would call this genre 'mediocre', and may explain why I had such a hard time with the sexism thing. Internal rhyming / assonance and other stuff could be done by rewarding the model for using it during the beam search. I suspect it's going to be hard to tune how to set the reward so it still makes sense, it only just makes sense already. I also think at this point you would really need to sit down and figure out what rhymes and what doesn't. I did the quickest, hackiest thing I could think of, which was good enough for generating end rhymes, but if you want to have internal rhymes and not massively constrain the model, I think you need to work it out purely from analysing the syllables.

kastnerkyle 10y ago

OP and others may be interested in this approach by Sony CSL from ~2012 using constraint Markov chains for matching lyrical style with explicit rhyme constraints, including (e.g.) Bob Dylan style with constraint of Beatles' Yesterday [0]. This is the precursor to the work that brought on "Daddy's Car"[1], though the techniques appear to have changed a bit reading their recent publications. I find their paper on recent approaches for music synthesis with style matching or other constraints [2] quite readable, and really good! I am hoping to build out a version of what is described by the paper. There is also a very nice talk by F. Pachet on this subject here [3], and his older videos are good too (they should be in the recommendations of YouTube). Still reading about DeepRhyme, but it seems great. Also a fantastic writeup! [0] https://www.csl.sony.fr/downloads/papers/2012/barbieri-12a.pdf https://www.csl.sony.fr/downloads/papers/2012/barbieri-12a.p... [1] https://www.youtube.com/watch?v=LSHZ_b05W7o https://www.youtube.com/watch?v=LSHZ_b05W7o [2] https://arxiv.org/abs/1609.05152 https://arxiv.org/abs/1609.05152 [3] https://www.youtube.com/watch?v=5K4hn6cBUPU https://www.youtube.com/watch?v=5K4hn6cBUPU

sweezyjeezy 10y ago
Hey thanks! I will check this out tomorrow!

adamsea 10y ago

As a huge hip-hop head, I will just add that there are two aspects to evaluating rhymes: How it sounds when delivered by the artist, and how it reads. The best rap succeeds in both domains. Bob Dylan won the Nobel Prize because he succeeds in both domains. In this case, we obviously can only consider the second domain, how it reads. So we should evaluate the results exactly how we would evaluate any other poem (while giving due respect to the conventions and tropes of the form, ie, rap has its own classic metaphors and themes).

infinitone 10y ago

This is awesome! I've been meaning to do something like this for awhile.. great that someone else finally did it. Any chance you'd open source it?

casoninabox 10y ago

First, this is awesome. Great results. I have been working on exactly this for the past few weeks. I went ahead and produced a song and a lyric video from mine (I call it Lil' Markov). It started out because I noticed patterns in Lil Wayne song where he'd say some idiom and then rhyme it. It seemed so predictable I figured I could make a bot that automated Lil Wayne's process... From what I can tell our methods are very similar. I have a thing that allows you to input a concept and it will try to rap around that concept and a `whiteGuilt()` function (that's whats up with all the ninjas), but other than that our process is 90% similar. We should collaborate on something! Message me. Version 0: https://soundcloud.com/cason-clagg/aesopxcasonxmarkov https://soundcloud.com/cason-clagg/aesopxcasonxmarkov Version 1: https://www.youtube.com/watch?v=fmcrjH6BKag https://www.youtube.com/watch?v=fmcrjH6BKag Version 2: (No music yet, bit here's some sample lyrics based on the concept "beef") FIND MYSELF WITH MILLIONS THE PIGGIES THAT'LL SHARE HOPPIN' AROUND SATURN'S RING I'M JUICED OFF DOG DARE A DESCRIPTIVE LYRICIST LOUNGING IN LOVE LIKE YOU AIN'T EAGER TO KILL EACH ONE OF THE DOVES HE RAN FOR A BEAST HE'S A PARAGON OF YORE DIDN'T THINK OF HIS BLOOD ON THE FACTORY FLOOR

lifeformed 10y ago

Next, you should plug this into a text-to-speech program and play it with a simple beat generator.

grandalf 10y ago

It would be fascinating to try training it on some HRC and Trump speeches. I am doubtful most people would be able to tell the difference between generated remarks and the real thing.