DAdaptation doesn't work and causes TypeError

[alectprasad]

`- LoRA Config:

• Loading network module: lycoris.kohya
• network args: ['conv_dim=8', 'conv_alpha=1', 'algo=lora']
• lycoris.kohya linear_dim set to: 16
• lycoris.kohya linear_alpha set to: 8
• lycoris.kohya conv_dim set to: 8
• lycoris.kohya conv_alpha set to: 1
• No LoRA weight loaded.
  • Optimizer Config:
• Additional network category: LoCon_Lycoris
• Using DAdaptation as Optimizer
• Optimizer Args: --optimizer_args ["decouple=True","weight_decay=0.4","betas=0.9,0.99"]
• Train UNet and Text Encoder
  • UNet learning rate: 1
  • Text encoder learning rate: 1
• Learning rate warmup steps: 0.1
• Learning rate Scheduler: cosine`

the next step Training Config throws

╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ in <cell line: 40>:63                                                                            │
│ in <module>:1                                                                                    │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
TypeError: string indices must be integers