Skip to content

Latest commit

Β 

History

History
69 lines (64 loc) Β· 3.02 KB

TRANSLATION_GUIDE.md

File metadata and controls

69 lines (64 loc) Β· 3.02 KB

λ²ˆμ—­ κ·œμΉ™

첫번째둜 ν•΄λ‹Ή μš©μ–΄κ°€ μΆœν˜„ν•˜μ˜€μ„ λ•ŒλŠ” 'ν•œκΈ€(μ˜μ–΄)'둜 λ²ˆμ—­ν•˜κ³ , μ΄ν›„λΆ€ν„°λŠ” ν•œκΈ€λ‘œλ§Œ λ²ˆμ—­ν•©λ‹ˆλ‹€. (예. including transposing, indexing, ... => μ „μΉ˜(transposing), 인덱싱(indexing), ...)

μš©μ–΄ μ‚¬μš© κ·œμΉ™

  1. μ•„λž˜ μš©μ–΄κ°€ μ μ ˆν•˜λ©΄ μ•„λž˜ ν‘œμ˜ μš©μ–΄λ₯Ό μ‚¬μš©ν•©λ‹ˆλ‹€.
  2. μ§€μ •λœ μš©μ–΄κ°€ μ—†λ‹€λ©΄ μ•„λž˜ μ‚¬μ΄νŠΈλ₯Ό μ°Έκ³ ν•˜μ—¬ μ‚¬μš©ν•©λ‹ˆλ‹€.

μ—†μœΌλ©΄ μ μ ˆν•œ 단어λ₯Ό μ‚¬μš©ν•˜κ³ , μ•„λž˜ λͺ©λ‘μ— λ‚΄μš©μ„ μΆ”κ°€ν•©λ‹ˆλ‹€.

영문 ν•œκΈ€ μž‘μ„±μž μΆ”κ°€ μ„€λͺ…
Acknowledgements κ°μ‚¬μ˜ 말 λ°•μ •ν™˜
API endpoint API μ—”λ“œν¬μΈνŠΈ λ°•μ •ν™˜ 음차 ν‘œκΈ°
argument 인자 λ°•μ •ν™˜
Audio μ˜€λ””μ˜€ λ°•μ •ν™˜ ToC의 λΆ„λ₯˜λͺ…μž…λ‹ˆλ‹€.
autograd Autograd ν™©μ„±μˆ˜ λ²ˆμ—­μ•ˆν•¨
Batch Normalization 배치 μ •κ·œν™” λ°•μ •ν™˜
dataset 데이터셋 λ°•μ •ν™˜ 음차 ν‘œκΈ°
deep neural network 심측 신경망 λ°•μ •ν™˜
derivative λ„ν•¨μˆ˜ λ°•μ •ν™˜
Drop-out Drop-out ν™©μ„±μˆ˜ λ²ˆμ—­μ•ˆν•¨
epoch 에폭 λ°•μ •ν™˜ 음차 ν‘œκΈ°
evaluation mode 평가 λͺ¨λ“œ λ°•μ •ν™˜
feed data through model 데이터λ₯Ό λͺ¨λΈμ— 제곡
Feed-forward network μˆœμ „νŒŒ 신경망 λ°•μ •ν™˜
Generative 생성 λͺ¨λΈ λ°•μ •ν™˜ ToC의 λΆ„λ₯˜λͺ…μž…λ‹ˆλ‹€.
Getting Started tutorial μ‹œμž‘ν•˜κΈ° νŠœν† λ¦¬μ–Ό λ°•μ •ν™˜ ToC의 Getting Startedλ₯Ό λœ»ν•©λ‹ˆλ‹€.
gradient 변화도 λ°•μ •ν™˜
Image 이미지 λ°•μ •ν™˜ ToC의 λΆ„λ₯˜λͺ…μž…λ‹ˆλ‹€.
in-place 제자리 ν—ˆλ‚¨κ·œ
instance μΈμŠ€ν„΄μŠ€ λ°•μ •ν™˜ 음차 ν‘œκΈ°
instantiate μƒμ„±ν•˜λ‹€ λ°•μ •ν™˜
Layer 계측 λ°•μ •ν™˜
learning rate, lr ν•™μŠ΅λ₯  λ°•μ •ν™˜
loss 손싀 λ°•μ •ν™˜
matrix ν–‰λ ¬ λ°•μ •ν™˜
mean-squared error ν‰κ· μ œκ³±μ˜€μ°¨ ν—ˆλ‚¨κ·œ
MelScale MelScale
mini-batch λ―Έλ‹ˆ 배치 λ°•μ •ν™˜ 음차 ν‘œκΈ°
momentum λͺ¨λ©˜ν…€ λ°•μ •ν™˜ 음차 ν‘œκΈ°
normalize μ •κ·œν™” ν—ˆλ‚¨κ·œ
NumPy NumPy λ°•μ •ν™˜ λ²ˆμ—­ν•˜μ§€ μ•ŠμŒ
One-Hot One-Hot ν™©μ„±μˆ˜ λ²ˆμ—­μ•ˆν•¨
Optimizer μ˜΅ν‹°λ§ˆμ΄μ € λ°•μ •ν™˜ 음차 ν‘œκΈ°
output 좜λ ₯ λ°•μ •ν™˜
over-fitting 과적합 ν™©μ„±μˆ˜
parameter λ§€κ°œλ³€μˆ˜ λ°•μ •ν™˜
placeholder ν”Œλ ˆμ΄μŠ€ν™€λ” λ°•μ •ν™˜ 음차 ν‘œκΈ°
plotting 도식화 ν™©μ„±μˆ˜
Production (environment, use) Production ν—ˆλ‚¨κ·œ λ²ˆμ—­ν•˜μ§€ μ•ŠμŒ
rank 0 0-μˆœμœ„ λ°•μ •ν™˜
Read later 더 읽을거리 λ°•μ •ν™˜
recap μš”μ•½ λ°•μ •ν™˜
resample λ¦¬μƒ˜ν”Œ
resizing 크기 λ³€κ²½ λ°•μ •ν™˜
sampling rate μƒ˜ν”Œλ§ 레이트
scenario μ‹œλ‚˜λ¦¬μ˜€ λ°•μ •ν™˜ 음차 ν‘œκΈ°
shape shape ν—ˆλ‚¨κ·œ λ²ˆμ—­ν•˜μ§€ μ•ŠμŒ
size 크기 λ°•μ •ν™˜
Tensor / Tensors Tensor λ°•μ •ν™˜ λ²ˆμ—­ν•˜μ§€ μ•ŠμŒ
Text ν…μŠ€νŠΈ λ°•μ •ν™˜ ToC의 λΆ„λ₯˜λͺ…μž…λ‹ˆλ‹€.
track (computation) history μ—°μ‚° 기둝을 μΆ”μ ν•˜λ‹€ λ°•μ •ν™˜
warmstart λΉ λ₯΄κ²Œ μ‹œμž‘ν•˜κΈ° λ°•μ •ν™˜ Warmstarting Model = λΉ λ₯΄κ²Œ λͺ¨λΈ μ‹œμž‘ν•˜κΈ°
weight κ°€μ€‘μΉ˜ λ°•μ •ν™˜
wrapper 래퍼 λ°•μ •ν™˜ 음차 ν‘œκΈ°