Giant_Panda_Fu_Bao / README.md
kangting87's picture
Update README.md
4f8ee92 verified
|
raw
history blame
1.46 kB
---
license: cc-by-nc-nd-4.0
language:
- ko
- en
pipeline_tag: audio-to-audio
tags:
- animal
- music
---
<img src=https://huggingface.co/kangting87/Giant_Panda_Fu_Bao/resolve/main/Giant_Panda_(Fu_Bao).png>
<br>
<a href="https://huggingface.co/kangting87/Giant_Panda_Fu_Bao/resolve/main/Giant_Panda_(Fu_Bao)_10e_210s_best_epoch_v2.zip?download=true" target="_blank">https://huggingface.co/kangting87/Giant_Panda_Fu_Bao/resolve/main/Giant_Panda_(Fu_Bao)_10e_210s_best_epoch_v2.zip?download=true</a>
<br>
<br>
Technology: RVC
<br>
Extraction Type: rmvpe
<br>
Number of Epochs: 10
<br>
Steps: 210
<br>
Model Architecture: v2
<br>
Sampling Rate: 48k
<br>
Pitch Guidance: True
<br>
Dataset Length: 1 min and 59 seconds
<br>
Overtrain Info: Smoothed loss_g 8.659 and loss_d 3.483
<br>
<br>
λ³Έ AI λͺ¨λΈμ˜ μŒμ—­λŒ€λŠ” λŒ€λž΅ B2 μ—μ„œλΆ€ν„° D5 κΉŒμ§€μž…λ‹ˆλ‹€.
<br>
λ§Œμ•½ B2 보닀 μ•„λž˜λ‘œ λ‚΄λ €κ°€κ±°λ‚˜ ν˜Ήμ€ D5 보닀 μœ„λ‘œ μ˜¬λΌκ°€λ©΄, μ†Œλ¦¬κ°€ μ†μƒλ˜λŠ” ν˜„μƒμ΄ μ„œμ„œνžˆ λ°œμƒν•©λ‹ˆλ‹€.
<br>
(This AI model approximately has the vocal range of B2~D5, and if it goes below B2 or above D5, the sound damaging phenomenon gradually begins to occur.)
<br>
<br>
ν‘Έλ°”μ˜€ 울음 μ†Œλ¦¬ ν‘œλ³Έμ΄ λ§Žμ§€ μ•Šμ€ κ΄€κ³„λ‘œ, 에폭 (epoch) μˆ˜λ„ 맀우 μ μŠ΅λ‹ˆλ‹€.
<br>
(Because there are not many samples of Fu Bao's voices, the number of epochs is also quite small.)
<br>
<br>
상업적 μ΄μš©μ„ κΈˆμ§€ν•©λ‹ˆλ‹€.
<br>
(Commercial use is prohibited.)