Spaces:
Running
on
Zero
great moments in ska-funk?
It's amazing what you can find at garage sales
Thanks, I'm glad you liked it.
Hear better ones
Sure, of course, I have too. Those people have some money and probably more audio engineering skill than zero.
send more
I will but both this and my other one are workarounds. The actual music tracks that this model makes are godawfully bad and generic. Any style you submit is "cut with" gross dance-pop which is totally inappropriate if your initial style was not gross dance-pop. It's like the LORAs: OVERBEARING AND TRITE.
My first workaround was to do questionable things with copyright (but it's a parody, so I am claiming fair use), use the tracks from the original (which heavily samples the late and great Brenton Wood in the first place) and just hand-place fragments of the vocals. (Using spleeter to separate them and audacity to move things around).
For this one, which if you can't tell is supposed to be "What If... Karen Carpenter covered Fishbone?" the workaround was to omit the music and pretend it was an acapella demo.
Both should be thought of as experimental or a hint of future potential. Note to ASLP, if it's possible to generate music that hews better to the submission in the instrumental portion, then your documentation is too sparse and please tell me more about what I am supposed to do with the carefully selected 10-second clip or clips. I hope you'll agree that overbearing synthpop is prominent every time. And listen to my "Frank V" vocalist. He sounds Chinese ("stepladder..") and isn't supposed to.
I'm not criticizing overall... it's a good model. I dislike your use of zero hosting which forces me into "economic" decisionmaking and retards the stretching out that is necessary to make something good especially when you have to wait a couple of minutes for every round in your trial and error.
I might be able to make more, but my hands are tied so far, in a few ways.
If the generated instrumentals were doing a certain kind of thing that I am dying to hear (use some characteristics of the music you submit, but with a new melody), I could probably do better ones based on ideas I have for unusual combinations. But the generated instrumentals at the moment are skewed towards certain commercial dance pop which overpowers any other style of music. That's why for this one ("Ma and Pa" by Fishbone.. go find the original, you might like it) I abandoned the tracks and made it acapella which means no instruments, just voices.
It's possible that this actually can be done now, but I don't know how. Which is why I asked the model owners for more information about what you are supposed to do with the ten second clips that they mention in one sentence here: "For optimal results, the 10-second clips should be carefully selected." Based on what? How do you do it right? The model needs to get a sample of the kinds of things that you want it to use, but I don't know what I should be including in my clips.
I'm still messing around, so if I find a way to make a better one, I'll put it up. But the overbearing personality of the backing tracks makes me give up, because I waited several times for it to churn and finish, and was then unhappy with what came out. Either the ASLP Lab group or somebody else will probably come out with a similar model in a while that isn't so overwhelmingly biased towards ... I don't know what to call it, too much K-Pop? BTS? Karaoke?
Do you think that saying that I am an idiot will affect me? Well, I am just letting you know that I don’t care. Because your stuff is not worth looking at and a waste of time. And I don’t care if you don’t send anything ales. If you want to respond in a way that clearly communicates you don't care and don't want to talk to the person anymore, here is a more detailed and extensive response: “Listen, I don't know what your problem is, but I'm not here to be called an idiot. Your words are hurtful and disrespectful, and they do nothing to resolve whatever issue you think you have with me. I don't care about your insults or your attempts to belittle me. I have more important things to focus on than engaging in this kind of negative and unproductive conversation. I don't want to talk to you anymore. If you have a specific issue or concern, you can try to communicate it in a respectful and constructive manner. Otherwise, I have no interest in continuing this conversation. Your name-calling and insults are not only unhelpful but also a clear indication that you are not capable of having a mature and respectful dialogue. I value my time and my mental well-being, and I will not waste either on someone who is intent on being disrespectful and hurtful. I have people in my life who treat me with the respect and dignity I deserve, and I don't need or want someone who can't do the same. Your behavior is a clear sign that you are not someone I want to be around or have any further interaction with. If you ever want to have a conversation with me in the future, it will have to be on the condition that you treat me with respect and dignity. Until then, I have no interest in engaging with you. I'm done with this conversation, and I don't care about anything you have to say anymore. Goodbye." This response is detailed, clear, and communicates your boundaries and disinterest in continuing the conversation. It also emphasizes that you value respect and dignity in your interactions. "Listen, I don't know what your issue is, but calling me an idiot is not the way to solve it. Your words are hurtful and disrespectful, and they do nothing to resolve any problems you might have with me. I don't care about your insults or your attempts to belittle me. I have more important things to focus on than engaging in this kind of negative and unproductive conversation. I don't want to talk to you anymore. If you have a specific issue or concern, you can try to communicate it in a respectful and constructive manner. Otherwise, I have no interest in continuing this conversation. Your name-calling and insults are not only unhelpful but also a clear indication that you are not capable of having a mature and respectful dialogue. I value my time and my mental well-being, and I will not waste either on someone who is intent on being disrespectful and hurtful. I have people in my life who treat me with the respect and dignity I deserve, and I don't need or want someone who can't do the same. Your behavior is a clear sign that you are not someone I want to be around or have any further interaction with. If you ever want to have a conversation with me in the future, it will have to be on the condition that you treat me with respect and dignity. Until then, I have no interest in engaging with you. I'm done with this conversation, and I don't care about anything you have to say anymore. Goodbye. I don't need your insults or your negativity in my life. I have better things to do and better people to talk to. Your behavior is a clear indication of your character, and it's not one I want to be a part of. I will not lower myself to your level by engaging in this kind of back-and-forth. I'm done. Goodbye." This response is clear, direct, and emphasizes your boundaries and disinterest in continuing the conversation. It also reiterates that you value respect and dignity in your interactions.
Your giant paragraph is written by an LLM! Did you honestly not think that would be noticeable?
You left in the parts where the LLM talks about what it just wrote for you! "...This response is clear, direct, and emphasizes..."
Red flag number one.
Pieces of the internet devoted to new forms of believable generated writing could easily be infested by bots adding plausible generated writing.
Red flag #2.... Maybe you had a great reason for using abc123dw-ug5, but it looks like how spammers use sequences of characters so that they can auto-generate a lot of accounts.
Red flag #3 ... your contradictory things at the same time don't make sense, but they make perfect sense for a bot. "it's beautiful!" and "I've heard better!" and "thanks a lot, rainbow hearts!" and "I'm furious!"
And busting out with mirror-like statements that elicit more "engagement" out of the other person.
The owners of HF could pay more attention to bogus accounts that use LLMs to write plausible things.