Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -4,17 +4,23 @@ kudos to mrfakename for the base gradio code I'm borrowing here.
|
|
4 |
|
5 |
日本語用
|
6 |
|
|
|
|
|
7 |
Unfortunately, due to the variation in how floating-point operations are performed across different devices,
|
8 |
and given the intrinsic characteristics of models that incorporate diffusion components,
|
9 |
-
it is unlikely that you will achieve identical results to those obtained on
|
10 |
So, the output you're about to hear may not accurately reflect the true performance of the model.
|
11 |
|
|
|
|
|
12 |
=========
|
13 |
|
14 |
|
|
|
15 |
残念ながら、異なるデバイスで浮動小数点演算が異なる方法で行われるため、およびDiffusionコンポーネントを取り入れたモデルの固有の特性を考慮すると、
|
16 |
モデルが元々トレーニングされたデバイスで得られた結果と同じ結果を得ることは難しいでしょう。
|
17 |
-
|
|
|
18 |
|
19 |
**
|
20 |
"""
|
|
|
4 |
|
5 |
日本語用
|
6 |
|
7 |
+
You will probably experience slight artifacts at the beginning or at the end of the output, which is not there on my server.
|
8 |
+
|
9 |
Unfortunately, due to the variation in how floating-point operations are performed across different devices,
|
10 |
and given the intrinsic characteristics of models that incorporate diffusion components,
|
11 |
+
it is unlikely that you will achieve identical results to those obtained on my server, where the model was originally trained.
|
12 |
So, the output you're about to hear may not accurately reflect the true performance of the model.
|
13 |
|
14 |
+
|
15 |
+
|
16 |
=========
|
17 |
|
18 |
|
19 |
+
|
20 |
残念ながら、異なるデバイスで浮動小数点演算が異なる方法で行われるため、およびDiffusionコンポーネントを取り入れたモデルの固有の特性を考慮すると、
|
21 |
モデルが元々トレーニングされたデバイスで得られた結果と同じ結果を得ることは難しいでしょう。
|
22 |
+
その結果、以下で体験するパフォーマンスはモデルの真の性能を正確に反映していません。
|
23 |
+
そのため、音声の開始時または終了時にアーティファクトが発生する可能性があります。
|
24 |
|
25 |
**
|
26 |
"""
|