text stringlengths 7 228 |
|---|
Extreme__More Than Words |
Extreme__More Than Words.4 |
Extreme__More Than Words.7 |
Extreme__More Than Words.2 |
Extreme__More Than Words.5 |
Extreme__More Than Words.1 |
Extreme__More Than Words.3 |
Extreme__More Than Words.6 |
Fatboy Slim__Right Here, Right Now |
Fatboy Slim__Right Here, Right Now.1 |
Fatboy Slim__Praise You |
Fatboy Slim__Going Out of My Head |
Fatboy Slim__Praise You.1 |
Fatboy Slim__How Can You Hear Us? |
The Clash__Rock in the Casbah.1 |
The Clash__London's Burning |
The Clash__Rock in the Casbah |
The Clash__Rock in the Casbah.2 |
The Clash__Should I Stay or Should I Go |
The Clash__Should I Stay or Should I Go.1 |
Mily Balakirev__Islamey |
Buena Vista Social Club__El cuarto de Tula |
Buena Vista Social Club__Murmullo |
Buena Vista Social Club__Candela |
No Doubt__Don't Speak.3 |
No Doubt__Don't Speak.2 |
No Doubt__Just a Girl.3 |
No Doubt__Don't Speak.5 |
No Doubt__Don't Speak.1 |
No Doubt__Don't Speak.4 |
No Doubt__Hey You |
No Doubt__Just a Girl |
No Doubt__Just a Girl.2 |
No Doubt__Don't Speak |
No Doubt__Ex-Girlfriend |
No Doubt__Just a Girl.4 |
No Doubt__Just a Girl.1 |
No Doubt__Spiderwebs |
No Doubt__Don't Speak.6 |
Izabella__Shame Shame Shame |
KWS__Please Don't Go |
Bizet Georges__Farandole (from |
Bizet Georges__Carmen |
Bizet Georges__Carmen Highlights - Habanera |
Renee de Haan__Vuile huichelaar |
Renee de Haan__De nacht is nog zo lang |
Johnson James P__Carolina Shout |
Vanessa Paradis__Be My Baby |
Nik Kershaw__Wouldn't It Be Good |
Terry Jacks__Seasons in the Sun.2 |
Terry Jacks__Seasons in the Sun.1 |
Terry Jacks__Seasons in the Sun |
The Shadows__FBI |
The Shadows__Walk Don't Run |
The Shadows__Wonderful Land |
The Shadows__Dance On |
The Shadows__Sleepwalk.2 |
The Shadows__Kontiki |
The Shadows__Atlantis.1 |
The Shadows__Unchained Melody |
The Shadows__Guitar Tango |
The Shadows__Theme From 'The Deerhunter' (Cavatina) |
The Shadows__Apache.2 |
The Shadows__Atlantis |
The Shadows__Guitar Tango.1 |
The Shadows__Kontiki.1 |
The Shadows__Kontiki.2 |
The Shadows__Sleepwalk |
The Shadows__Riders in the Sky.1 |
The Shadows__Wonderful Land.1 |
The Shadows__Apache.3 |
The Shadows__Sleepwalk.1 |
The Shadows__Apache.1 |
The Shadows__Riders in the Sky |
The Shadows__Foot Tapper |
The Shadows__Apache |
The Shadows__Quartermaster Stores |
The Shadows__FBI.1 |
The Shadows__Guitar Boogie |
elgar__Pomp and Circumstances March No.1 |
Simon Paul__50 Ways to Leave Your Lover.1 |
Simon Paul__Loves Me Like a Rock |
Simon Paul__You Can Call Me Al.1 |
Simon Paul__At the Zoo |
Simon Paul__Hazy Shade of Winter |
Simon Paul__Late In The Evening |
Simon Paul__50 Ways to Leave Your Lover |
Simon Paul__Mother and Child Reunion |
Simon Paul__You Can Call Me Al |
Simon Paul__Me and Julio Down by the School Yard |
Simon Paul__Kodachrome |
The Ventures__Joker's Wild |
The Ventures__Bulldog |
The Ventures__Bluebird |
The Ventures__Walk Don't Run.1 |
The Ventures__Bird Rockers |
The Ventures__Walk Don't Run |
The Ventures__I Feel Fine |
The Ventures__Besame Mucho |
The Ventures__Gringo |
End of preview. Expand
in Data Studio
LMD Deduplication Supplements
This repository provides pre-computed embedding files extracted from the Lakh MIDI Dataset (LMD-clean and LMD-full) using the CAugBERT and CLaMP-1024 models.
These embeddings were used in our paper:
"On the De-duplication of the Lakh MIDI Dataset" (ISMIR 2025)
[Paper] | [GitHub Code]
Contents
Each folder includes:
embeddings.pt: Torch tensor of embeddings (shape: N × D)refs.txt: List of MIDI filenames corresponding to each embedding row
Usage
import torch
# Load embeddings
embeddings_dir = './lmd_full_to_lmd_full_0.95__caugbert_embedding.npy/'
emb = torch.load(embeddings_dir + "embeddings.pt")
# Load references (MIDI filenames)
with open(embeddings_dir + "refs.txt") as f:
refs = [line.strip() for line in f]
print(emb.shape, len(refs))
# Example: torch.Size([168662, 512]) 168662
For details on how these embeddings are used in duplicate detection and evaluation, please refer to the
Evaluate and De-duplication sections of the main repository.
Note
Due to storage limits, we only provide embeddings from CAugBERT and CLaMP-1024 here. If you need embeddings from other models (e.g., CLaMP-512, MusicBERT, etc.), please contact: Eunjin Choi (jech@kaist.ac.kr)
- Downloads last month
- 20