Datasets:

eduhk-lin3046
/

Group6_Eminem_dataset

Modalities:

Dataset Preview

Split (1)

train

The full dataset viewer is not available (click to read why). Only showing a preview of the rows.

The dataset generation failed

Error code:   DatasetGenerationError
Exception:    UnicodeDecodeError
Message:      'utf-8' codec can't decode byte 0xd1 in position 3022: invalid continuation byte
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1855, in _prepare_split_single
                  for _, table in generator:
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/packaged_modules/text/text.py", line 73, in _generate_tables
                  batch = f.read(self.config.chunksize)
                File "/usr/local/lib/python3.9/codecs.py", line 322, in decode
                  (result, consumed) = self._buffer_decode(data, self.errors, final)
              UnicodeDecodeError: 'utf-8' codec can't decode byte 0xd1 in position 3022: invalid continuation byte
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1438, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1050, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 925, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1001, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1742, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1898, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

text string
when i was just a little baby boy,
my momma used to tell me these crazy things
she used to tell me my daddy was an evil man,
she used to tell me he hated me

but then i got a little bit older
and i realized, she was the crazy one
but there was nothing i could do or say to try to change it
'cause that's just the way she was

they said i can't rap about being broke no more
they ain't say i can't rap about coke no more
slut, you think i won't choke no whore
'til the vocal cords don't work in her throat no more?!

these motherfuckers are thinking i'm playing
thinking i'm saying the shit 'cause i'm thinking it just to be saying it
put your hands down bitch, i ain't gon' shoot you
i'mma pull you to this bullet, and put it through you

shut up slut, you're causing too much chaos
just bend over and take it like a slut, ok ma?
"oh, now he's raping his own mother,
abusing a whore, snorting coke, and we gave him the rolling stone cover?"

you're goddamn right bitch, and now it's too late
i'm triple platinum and tragedies happen in two states
i invented violence, you vile venomous volatile bitches
vain vicodin, vrinn vrinn, vrinnn!

texas chainsaw, left his brains all
dangling from his neck, while his head barely hangs on
blood, guts, guns, cuts
knives, lives, wives, nuns, sluts

bitch i'mma kill you! you don't wanna fuck with me
girls neither – you ain't nothing but a slut to me
bitch i'mma kill you! you ain't got the balls to beef
we ain't gon' never stop beefing i don't squash the beef

you better kill me! i'mma be another rapper dead
for popping off at the mouth with shit i shouldn't have said
but when they kill me – i'm bringing the world with me
bitches too! you ain't nothing but a girl to me

i said you don't wanna fuck with shady ('cause why?)
'cause shady will fucking kill you
i said you don't wanna fuck with shady (why?)
'cause shady will fucking kill you

bitch i'mma kill you!
like a murder weapon, i'mma conceal you
in a closet with mildew, sheets, pillows and film you
fuck with me, i been through hell, shut the hell up!
i'm tryna develop these pictures of the devil to sell 'em

i ain't "acid rap", but i rap on acid
got a new blow-up doll and just had a strap-on added
whoops! is that a subliminal hint? no!
just criminal intent to sodomize women again
eminem offend? no! eminem'll insult
and if you ever give in to him, you give him an impulse
to do it again, then, if he does it again
you'll probably end up jumping out of something up on the 10th

bitch i'mma kill you, i ain't done this ain't the chorus
i ain't even drug you in the woods yet to paint the forest
a bloodstain is orange after you wash it three or four times in a tub
but that's normal ain't it norman?

serial killer hiding murder material
in a cereal box on top of your stereo
here we go again, we're out of our medicine
out of our minds, and we want in yours, let us in

or i'mma kill you! you don't wanna fuck with me
girls neither – you ain't nothing but a slut to me
bitch i'mma kill you! you ain't got the balls to beef
we ain't gon' never stop beefing i don't squash the beef

you better kill me! i'mma be another rapper dead
for popping off at the mouth with shit i shouldn't have said
but when they kill me i'm bringing the world with me
bitches too! you ain't nothing but a girl to me

i said you don't wanna fuck with shady ('cause why?)
'cause shady will fucking kill you
i said you don't wanna fuck with shady (why?)
'cause shady will fucking kill you

eh-heh, know why i say these things?
'cause ladies' screams keep creeping in shady's dreams
and the way things seem, i shouldn't have to pay these shrinks
this eighty gs a week to say the same things tweece!

twice? whatever, i hate these things
fuck shots! i hope the weed'll outweigh these drinks
motherfuckers want me to come on their radio shows
just to argue with 'em cause their ratings stink?

End of preview.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

LIN3046 Group Project: Eminem Rap Lyrics Dataset for AI Music Generation

Abstract

This project creates a curated dataset of Eminem’s rap lyrics from his 2000 album The Marshall Mathers LP to enhance AI-driven music generation. The dataset includes annotated lyrics, metadata, and audio tracks, with a focus on emotional depth (Anger, Sadness, Satire, Pride, Love, Humor) and structural elements like rhyme schemes. Data was collected manually and programmatically, processed using TF-IDF analysis for keyword extraction, and annotated through a hybrid of zero-shot AI models and manual validation. Designed for researchers, this dataset aims to improve AI models' ability to generate culturally resonant and emotionally nuanced rap lyrics, bridging gaps in current text-only, generalized music generation frameworks. Key limitations include subjectivity in annotations and the absence of audio features like rhythm.

Dataset Overview

Dataset Name: Eminem_rap_dataset_18042025
Structure:

Lyrics (.txt): Cleaned lyrical content.
Annotated_Lyrics (.json): Emotion labels, rhyme patterns, and cultural references.
Metadata (.txt): Song details (timestamp, producer, etc.).
Songs (.wav): Audio tracks from YouTube, converted programmatically.
ARCHIVE: Raw and intermediate files.

Key Features:

Emotion Tags: 6 categories (Anger, Sadness, Satire, Pride, Love, Humor).
Rhyme Annotation: End-of-line rhymes identified using the pronouncing library.
Cultural References: Mentions of people, events, or locations.

Size: 70 minutes (1.2 hours) of audio, 5+ annotated songs.

Data Collection & Annotation

Lyrics Extraction:
- Lyrics were transcribed manually, excluding timestamps, sound effects, and non-lyrical elements.
- Converted to lowercase and tokenized for consistency.
Audio Collection:
- A Python script using YT-DLP downloaded and converted YouTube tracks to .wav format.
TF-IDF Analysis:
- Identified top keywords (e.g., "shady," "fuck") to guide emotion labeling.
Annotation Process:
- Sentiment: Zero-shot AI models + manual validation for emotional tags.
- Rhymes: Grouped by phonetic patterns (e.g., "A," "B").
- Cultural Context: Regex searches and manual verification.

Usage Recommendations

Style Mimicry: Train models to generate lyrics mirroring Eminem’s thematic and rhythmic patterns.
Emotional Depth: Leverage sentiment tags to link language patterns with specific emotions.
Cultural Relevance: Use annotated cultural references to enhance contextual authenticity.

Limitations

Subjectivity: Emotion and rhyme annotations may vary between annotators.
Audio Exclusion: Lacks beat, flow, or delivery data critical to rap.
Simplified Emotions: Six emotion labels may oversimplify complex lyrical sentiment.
Static Content: Does not capture evolving slang or collaborative artistry (e.g., freestyling).

References

For methodology and context, see the full report and citations in References section.

license: mit

Downloads last month: 5

Total file size:

814 MB