Hullwave · Gokstad AI Music Generator

Text-to-Music

Audio Duration (seconds)

30 s · Random (model-chosen 30-240, sends -1)

Enable Audio2Audio

Uses a reference audio file as a starting point.

LoRA Name or Path

Reference Audio (for Audio2Audio)

Upload an audio file to transform. (upload wiring TBD)

Reference Audio Strength

0.50

Support tags, descriptions, and scene. Use commas to separate different tags.
Tags and lyrics examples are from AI music generation community.

Preset

Tags

Support lyric structure tags like [verse], [chorus], [bridge].
Use [instrumental] or [inst] to generate instrumental music.

Lyrics

[verse]
In the quiet of my room, I'm searching for a tune,
Something fresh, something new, to light up the gloom.
Through the speakers, a melody, whispers in the night,
A symphony of sounds, igniting a spark of light.
                            
[chorus]
Discovering new music, in every beat I find,
A journey through the rhythm, a treasure to unwind.
In every note, a story, in every chord, a dream,
Discovering new music, in this world of sound and steam.
                            
[verse]
From the depths of the internet, a playlist appears,
A mix of the familiar, and the ones I've never heard.
I press play, and the world fades away,
In this sonic adventure, I'm on my way.
                            
[bridge]
From classical to punk, from jazz to indie rock,
Each genre a door, waiting to be unlocked.
With every song, I feel alive, with every beat, I'm free,
Discovering new music, as wild and as free as the sea.
                            
[chorus]
Discovering new music, in every beat I find,
A journey through the rhythm, a treasure to unwind.
In every note, a story, in every chord, a dream,
Discovering new music, in this world of sound and steam.

Settings

Basic Settings

Infer Steps

60 steps

Guidance Scale

15.0

When guidance_scale_lyric > 1 and guidance_scale_text > 1, this global guidance is ignored.

Guidance Scale (Text)

0.0

Only applies in double-condition CFG. Start with 5.0 for text, 1.5 for lyrics.

Guidance Scale (Lyric)

0.0

Manual Seeds (optional)

Comma-separated list or empty for random.

Advanced Settings

Scheduler Type

euler heun

euler is recommended; heun is slower.

CFG Type

cfg apg cfg_star

Entropy Rectifying Guidance (ERG)

use ERG for tag use ERG for lyric use ERG for diffusion

Granularity Scale (omega_scale)

10.0

Higher values can reduce artifacts; default is 10.

Guidance Interval

0.50

0.5 ≈ apply guidance in the middle half of the steps.

Guidance Interval Decay

0.00

If > 0, guidance scale decays toward Min Guidance Scale.

Min Guidance Scale

3.0

OSS Steps

Optional comma-separated step indices. Leave blank for standard sampling.

Variance

0.20

Retake Seeds (optional)

Variance

0.20

Repaint Seeds (optional)

Repaint Start Time

Repaint End Time

Repaint Source

text2music last_repaint upload

Edit Tags

Edit Lyrics

Edit Type

only_lyrics remix

edit_n_min

0.60

edit_n_max

1.00

Edit Source

text2music last_edit upload

Edit Seeds (optional)

Extend Seeds (optional)

Left Extend Length (seconds)

Right Extend Length (seconds)

Extend Source

text2music last_extend upload

Text2Music Parameters

Shows the last Text2Music payload/params used. (Generate something first.)

{}

Library

Gallery

Loading clips…

Output

Workspace

Text-to-Music

Settings

Library