black-eyed-peas-v1-crafted-variable-prompt-16-epochs-text-only-piano-prompts

This model is a fine-tuned version of facebook/musicgen-melody-large on the SWEET-DREAMBOOTHS/BLACK-EYED-PEAS-V1-PIANO-PROMPTS - DEFAULT dataset. It achieves the following results on the evaluation set:

• Loss: 3.6533
• Text Cosine Sim: 0.2462
• Audio Cosine Sim: 0.5594

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 456
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Text Cosine Sim	Audio Cosine Sim
7.791	4.9485	30	3.1004	0.3762	0.5584
6.6699	9.8969	60	3.2356	0.3653	0.6576
6.4394	14.8454	90	3.6373	0.2610	0.5900

Framework versions

• PEFT 0.10.0
• Transformers 4.41.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.18.1.dev0
• Tokenizers 0.19.1