Finetuning Stable Audio Open

>> YOUR LINK HERE: ___ http://youtube.com/watch?v=ex4OBD_lrds

00:00 - Getting started • 03:18 - Installing libraries model download • 20:31 - Switch to Jupyter Notebook with A6000s • 26:41 - Dataset upload organization • 40:10 - model_config.json params + math • 41:51 - Choosing demo prompts • 45:11 - Dataset configs • 52:56 - Crafting the train_py command • 58:06 - Start the training run • 01:00:38 - Wandb monitoring • 01:05:37 - Fixing get_custom_metadata • 01:17:15 - Restart the training run • 01:23:48 - Diffusion loss curves explained • 01:31:23 - 1st demo results • 01:42:09 - 2nd demo results • 01:52:44 - 3rd demo results • 02:03:29 - 4th demo results • 02:17:11 - 5th demo results • 02:20:21 - Unwrapping the checkpoint • 02:24:56 - Exporting the finetuned model • 02:31:10 - Running Gradio generating samples • 02:41:23 - Batch size experiments • 02:59:19 - Q A • 03:10:42 - Live track building with finetune samples • ty ‪@youngphlo‬

#############################

New on site