Voice Models v2
EasyAIVoice ↗
32,018 Verified Open Models 28,884 Multi-Pitch Audition Ready 100% Direct Source URLs (HF · Drive · MEGA)
j-hope of BTS
RVC v2 ·304 Epochs

j-hope of BTS

KLM 4.4 MRF-HIFIGAN 40K

RVC v2 304 Epochs Added Jan 23, 2025
Open Model Weights Vault · Hosted on Hugging Face · 174.1 MB
✓ Verified Direct URL
https://huggingface.co/natanworkspace/j-hope_KLM_4.4_40K_MRF-HIFIGAN/resolve/main/j-hope_304e_mrf-hifigan.zip?download=true
Download Raw Weights (.zip) ↗
Convert Audio with This Voice in EasyAIVoice
Carries this exact model URL and your selected pitch (0 st) into the cloud studio.
Zero GPU Setup
Audio Decision Loop · Sample Deck

Audition Across 18 Pitch References

Click any semitone cell below to hear how this voice performs on female and male reference vocals and lock your preferred pitch.

18 Reference Previews
FEMALE reference · Pitch 0 st
Generated reference preview · Ready to audition
Semitone Shift (`-16` to `+16` st) · Bar height indicates NISQA clarity score
Female Ref
Male Ref
Audible Female Audible Male
Raw model title: j-hope of BTS (KLM 4.4 MRF-HIFIGAN 40K) (304 Epochs)
CLAP Semantic Analysis · Voice Accords

Sonic Profile at Selected Pitch

Acoustic traits measured directly from the generated output sample. Switch pitches in the Sample Deck to see how timbre shifts.

Technical Profile

Architecture: RVC v2 · Training: 304 Epochs · Source Host: Hugging Face

Auditory Discovery · Close Alternatives

Voices That Sound Like j-hope of BTS

Close but not quite the right fit? Audition nearby candidates matched by vocal embedding and timbre.

High-clarity female voice
MALE · 0 st 5.13 MOS HF
High-clarity female voice
FEMALE · 0 st 5.12 MOS Direct
High-clarity female voice
FEMALE · 0 st 5.11 MOS HF
Libby Rue Better Nate Than Ever
High-clarity female voice
FEMALE · 0 st 5.10 MOS HF
Gallica Metaphor ReFantazio
High-clarity female voice
FEMALE · 0 st 5.10 MOS HF
High-clarity female voice
MALE · 0 st 5.09 MOS HF
Model Documentation & Notes

About j-hope of BTS (KLM 4.4 MRF-HIFIGAN 40K) (304 Epochs)

⚠️ This model is trained on a different vocoder (MRF-HIFIGAN) and is ONLY recommended to use this model on Applio (main branch). It won't work on any older forks (like mainline, mangio, etc.) but Applio. fp32 RMVPE 22 Minutes of datasets (rap and singing with 8:2 ratio) https://huggingface.co/natanworkspace/j-hope_KLM_4.4_40K_MRF-HIFIGAN/resolve/main/j-hope_304e_mrf-hifigan.zip?download=true

Provenance & Open Access
Catalog ID:#31285
Weights Host:Hugging Face
Archive Size:174.1 MB
Rights & Safety: Report / Takedown