Update readme to recommend finetuning config

Author: ZDisket

examples/multiband_pwgan/README.md

@@ -68,7 +68,7 @@ Download generator weights
 CUDA_VISIBLE_DEVICES=0 python examples/multiband_pwgan/train_multiband_pwgan.py \
   --train-dir ./dump/train/ \
   --dev-dir ./dump/valid/ \
-  --outdir ./examples/multiband_pwgan/exp/train.multiband_melgan.v1/ \
+  --outdir ./examples/multiband_pwgan/exp/train.multiband_melgan.v1ft/ \
   --config ./examples/multiband_pwgan/conf/multiband_pwgan.v1.yaml \
   --use-norm 1 \
   --generator_mixed_precision 1 \
@@ -94,4 +94,4 @@ Here is a learning curves of melgan based on this config [`multiband_melgan.v1.y
 
 1. https://github.com/kan-bayashi/ParallelWaveGAN
 2. [Parallel WaveGAN: A fast waveform generation model based on generative adversarial networks with multi-resolution spectrogram](https://arxiv.org/abs/1910.11480)
-3. [Multi-band MelGAN: Faster Waveform Generation for High-Quality Text-to-Speech](https://arxiv.org/abs/2005.05106)
+3. [Multi-band MelGAN: Faster Waveform Generation for High-Quality Text-to-Speech](https://arxiv.org/abs/2005.05106)