MusicLang – a controllable model for symbolic music gen (LLAMA2)

Hacker News

Hi HN, MusicLang confounder here ! After 18 months of R&D, we've just released our project to the open source community. I wanted to share it with you too, hoping it might be helpful for your projects. We started MusicLang with the aim of addressing the limitations of existing language models in music composition. Despite the increasing popularity of LLMs, none have fully met the requirements for music generation. Similarly, existing generative models for music mostly focus on audio and lack controllability over the output generated. This poor UX was the reason why we built MusicLang, which offers advanced control features over symbolic music generation. MusicLang is ideal if you: - Want to generate music exportable to your favorite DAW in MIDI format; - Seek control over chord progressions in the generated music; - Need to create variations or extend existing music; - Require fast performance on a laptop without GPU. MusicLang comprises several components: 1.MusicLang Foundation model: Our foundational model for creating and generating original MIDI soundtracks [0]; 2. MusicLang predict: Our AI prediction API for the MusicLang package [1]; 3. MusicLang Language: A new language for tonal music, enabling composers to load, write, transform, and predict symbolic music easily [2]. Instead of a long explanation, you can try MusicLang on our colab [3], explore our demo space [4], or our HuggingFace space [5]. For more technical details, visit our technical blog [6]. Mervin Prison also made a cool video showcasing our features [7]. We’re excited to hear your thoughts and feedback, this will help us a lot. Thanks :-) [0] MusicLang Foundation model: https://huggingface.co/musiclang/musiclang-v2 [1] MusicLang predict: https://github.com/musiclang/musiclang_predict?tab=readme-ov... [2] MusicLang language: https://github.com/MusicLang/musiclang [3] Colab: https://colab.research.google.com/drive/1MA2mek826c05BjbWk2n... [4] Demo space: https://huggingface.co/spaces/musiclang/musiclang-predict [5] HuggingFace space: https://huggingface.co/musiclang [6] Technical blog: https://musiclang.github.io/ [7] Mervin Praison’s video: https://www.youtube.com/watch?v=5lWgP4F5Q-o

Visit product View on Hacker NewsMar 17, 2024

Share card

Actual performance

2points

Did not reach leaderboard

Launch Intel predictions

Analyze your own launch →

Indie HackersFits the IH revenue-focused audience · Strong signals: started · Missing: supports, reddit linkedin, podcasting

91%91% predicted probability of success on Indie Hackers, based on ML models trained on real launch data.

best fitHighest predicted score across all platforms for this description.

Product HuntOn track for Day 1 leaderboard · Strong signals: model, google, new · Missing: mac, agents, macos

89%89% predicted probability of success on Product Hunt, based on ML models trained on real launch data.

Hacker NewsStrong engagement from HN community · Strong signals: excited, just released, exist · Missing: https docs, lua, filesystem

85%85% predicted probability of success on Hacker News, based on ML models trained on real launch data.

nativeThis product was originally launched on this platform.

TrustMRRLess likely to generate early MRR · Strong signals: video, month, google · Missing: mobile apps, ios, personal

46%46% predicted probability of success on TrustMRR, based on ML models trained on real launch data.

AppSumoMay struggle as an AppSumo deal · Missing: plus, platform, intuitive

44%44% predicted probability of success on AppSumo, based on ML models trained on real launch data.

Acquire.comPre-revenue stage for this audience · Missing: arr, mrr, revenue

16%16% predicted probability of success on Acquire.com, based on ML models trained on real launch data.

BetaListMay not resonate with beta-testers · Strong signals: audio · Missing: web3, chat, crypto

0%0% predicted probability of success on BetaList, based on ML models trained on real launch data.

Incorrect prediction on native model

Similar products

Erlang's Gen-Fsm in Go65%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Erlang's Gen-Fsm in Go

Hacker News5

Crowdraising52%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Next-Gen Crowdsourcing

Indie Hackers1$500/mob2b

Fidl52%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

The next-gen ticketing system.

Indie Hackers

Wozi40%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Next-gen sweetener:not only sweeten but also enhances flavor

Indie Hackers1b2b

Vermony60%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Automatic music gen

Indie Hackers1entertainment

Splash Pro (Gen-2) – Midjourney for music38%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Splash Pro (Gen-2) – Midjourney for music

Hacker News3

extremr: what's the loudest, most danceable, most energetic music?53%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

extremr: what's the loudest, most danceable, most energetic music?

Hacker News2

Music for Geeks and Nerds47%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Music for Geeks and Nerds

Hacker News221

Bangarang Boomerang - a music hack48%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Bangarang Boomerang - a music hack

Hacker News3

See your music with HTML560%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

See your music with HTML5

Hacker News7

Browse all products Analyze your launch