MiniMax Music 3 identity tournament
A controlled singer-identity comparison across Serj-only and Daron-only MiniMax Music 3
language-model LoRAs. It also compares random 128-frame training windows against continuation
windows while holding the dataset split, instrumental regularisation, optimizer, prompts,
inference seed, and LoRA-strength sweep fixed.
This card cross-links 640 FLAC renders hosted by the four source
repositories. Audio is not duplicated here.
Design
| leg | singer tags | window objective | instrumental regularisation | checkpoints |
|---|
| Serj, random window | Serj Tankian | random 128-frame crop | 50% singer-matched | 10 |
| Daron, random window | Daron Malakian | random 128-frame crop | 50% singer-matched | 10 |
| Serj, continuation window | Serj Tankian | continuation, 128-frame target | 50% singer-matched | 10 |
| Daron, continuation window | Daron Malakian | continuation, 128-frame target | 50% singer-matched | 10 |
Singer identity and training-window comparison (640 renders)
Serj-only and Daron-only LoRAs trained with random or continuation windows, each with singer-matched instrumental regularisation.
All renders use a 40-second maximum duration, 30 flow steps, flow CFG 1.7, and seed 20260824. MiniMax Music 3 may end a song before the duration cap.
Training legs
No trigger word / off-genre descriptors
Caption: Alternative pop-jazz, 102 BPM, brushed drums, upright bass, clean electric piano, airy male vocals, smooth verse-chorus songwriting, restrained dynamics, polished intimate mix.
LoRA strength 0.25
LoRA strength 0.50
LoRA strength 0.75
LoRA strength 1.00
Trigger word only
Caption: system of a down, soad style music
LoRA strength 0.25
LoRA strength 0.50
LoRA strength 0.75
LoRA strength 1.00
Matching descriptors, no trigger word
Caption: Lilting hypnotic arpeggios, a gentle melodic mid-tempo groove with layered harmonies, sudden explosive distorted riffs, unstable meter changes, theatrical male vocals, tightly syncopated bass and drums, abrupt shifts between fragile verses and frantic heavy choruses.
LoRA strength 0.25
LoRA strength 0.50
LoRA strength 0.75
LoRA strength 1.00
Trigger word + matching descriptors
Caption: system of a down, soad style music, lilting hypnotic arpeggios, a gentle melodic mid-tempo groove with layered harmonies, sudden explosive distorted riffs, unstable meter changes, theatrical male vocals, tightly syncopated bass and drums, abrupt shifts between fragile verses and frantic heavy choruses.
LoRA strength 0.25
LoRA strength 0.50
LoRA strength 0.75
LoRA strength 1.00
Evaluation lyrics
The lexical content is fixed; only the singer assignment in section tags changes.
Show Serj Tankian section tags
1[Intro]
2Count the teeth in the traffic light
3Paint the silence ultraviolet
4
5[Verse 1: Serj Tankian]
6Paper crowns on the factory floor
7Everybody wants a little more
8Wind-up prophets in a checkout line
9Selling borrowed seconds back as time
10
11[Pre-Chorus: Serj Tankian]
12Turn the dial, divide the sky
13Ask the wires to tell us why
14
15[Chorus: Serj Tankian]
16Wake up the walls, let the ceiling shake
17We built a feast from a plastic lake
18Hands in the air while the engines stall
19Who signed the order to swallow it all?
20
21[Verse 2: Serj Tankian]
22Velvet sirens in the avenue
23Paint every warning a pleasant blue
24I fed my shadow to the voting machine
25It came back polished, quiet, and clean
26
27[Bridge: Serj Tankian]
28Softly, softly, count to four
29Kick down the arithmetic door
30One for the hunger, two for the show
31Three for the ones who already know
32
33[Final Chorus: Serj Tankian]
34Wake up the walls, let the ceiling shake
35We built a feast from a plastic lake
36Hands in the air while the engines stall
37Who signed the order to swallow it all?
38
39[Outro]
40Count the teeth
41Kill the light
42Carry the silence
43Into the night
Show Daron Malakian section tags
1[Intro]
2Count the teeth in the traffic light
3Paint the silence ultraviolet
4
5[Verse 1: Daron Malakian]
6Paper crowns on the factory floor
7Everybody wants a little more
8Wind-up prophets in a checkout line
9Selling borrowed seconds back as time
10
11[Pre-Chorus: Daron Malakian]
12Turn the dial, divide the sky
13Ask the wires to tell us why
14
15[Chorus: Daron Malakian]
16Wake up the walls, let the ceiling shake
17We built a feast from a plastic lake
18Hands in the air while the engines stall
19Who signed the order to swallow it all?
20
21[Verse 2: Daron Malakian]
22Velvet sirens in the avenue
23Paint every warning a pleasant blue
24I fed my shadow to the voting machine
25It came back polished, quiet, and clean
26
27[Bridge: Daron Malakian]
28Softly, softly, count to four
29Kick down the arithmetic door
30One for the hunger, two for the show
31Three for the ones who already know
32
33[Final Chorus: Daron Malakian]
34Wake up the walls, let the ceiling shake
35We built a feast from a plastic lake
36Hands in the air while the engines stall
37Who signed the order to swallow it all?
38
39[Outro]
40Count the teeth
41Kill the light
42Carry the silence
43Into the night
Inference notes
Each set of four LoRA strengths was generated in one batched pass with a shared random stream.
NextLat predictor and XM route-embedding sidecars are training-only and are not used at inference.
MiniMax Music 3 may end a song before the configured duration cap.
