Make sample rate in finetune wiki a bit more clear

[RenNagasaki]

Hey @erew123 you may want to edit the part of the finetuning wiki where you explain the sample rate.
While it's true that any sample rate works but I had another run with my (pretty darn big, 29000 Files) dataset the last few days while creating the vast.ai template.

You can find what I mean here: Link)
Point "2. Format Requirements" -> "Sample Rate"

Both tries were run with the same settings on a 4090 + 100 GB Ram: Vanilla Settings (Batch Size 4, Gradient 1, Max Audio 20, AdamW, and so on)

• Average time for 50 steps with my original sample rate audio (44100) -> 15 Minutes
• Average time for 50 steps with pre-converted audio (22050) -> 40 Seconds

That's and improvement of around 95%.

Should be noted in my opinion.

Kind regards,
Ren