SMITIN: Self-Monitored Inference-Time INtervention for Generative Music Transformers
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917907909312512 |
|---|---|
| author | Koo, Junghyun Wichern, Gordon Germain, Francois G. Khurana, Sameer Roux, Jonathan Le |
| author_facet | Koo, Junghyun Wichern, Gordon Germain, Francois G. Khurana, Sameer Roux, Jonathan Le |
| contents | We introduce Self-Monitored Inference-Time INtervention (SMITIN), an approach for controlling an autoregressive generative music transformer using classifier probes. These simple logistic regression probes are trained on the output of each attention head in the transformer using a small dataset of audio examples both exhibiting and missing a specific musical trait (e.g., the presence/absence of drums, or real/synthetic music). We then steer the attention heads in the probe direction, ensuring the generative model output captures the desired musical trait. Additionally, we monitor the probe output to avoid adding an excessive amount of intervention into the autoregressive generation, which could lead to temporally incoherent music. We validate our results objectively and subjectively for both audio continuation and text-to-music applications, demonstrating the ability to add controls to large generative models for which retraining or even fine-tuning is impractical for most musicians.
Audio samples of the proposed intervention approach are available on our demo page http://tinyurl.com/smitin . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_02252 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | SMITIN: Self-Monitored Inference-Time INtervention for Generative Music Transformers Koo, Junghyun Wichern, Gordon Germain, Francois G. Khurana, Sameer Roux, Jonathan Le Sound Audio and Speech Processing We introduce Self-Monitored Inference-Time INtervention (SMITIN), an approach for controlling an autoregressive generative music transformer using classifier probes. These simple logistic regression probes are trained on the output of each attention head in the transformer using a small dataset of audio examples both exhibiting and missing a specific musical trait (e.g., the presence/absence of drums, or real/synthetic music). We then steer the attention heads in the probe direction, ensuring the generative model output captures the desired musical trait. Additionally, we monitor the probe output to avoid adding an excessive amount of intervention into the autoregressive generation, which could lead to temporally incoherent music. We validate our results objectively and subjectively for both audio continuation and text-to-music applications, demonstrating the ability to add controls to large generative models for which retraining or even fine-tuning is impractical for most musicians. Audio samples of the proposed intervention approach are available on our demo page http://tinyurl.com/smitin . |
| title | SMITIN: Self-Monitored Inference-Time INtervention for Generative Music Transformers |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2404.02252 |