File size: 1,859 Bytes
55a9e2d 043b71d 55a9e2d 043b71d 55a9e2d 043b71d 55a9e2d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
---
library_name: peft
license: cc-by-nc-4.0
base_model: facebook/musicgen-small
tags:
- text-to-audio
- taufiqsyed/salami_cleaned_sampled
- generated_from_trainer
model-index:
- name: salami_truncsplit_finetune_model_16
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# salami_truncsplit_finetune_model_16
This model is a fine-tuned version of [facebook/musicgen-small](https://huggingface.co/facebook/musicgen-small) on the TAUFIQSYED/SALAMI_CLEANED_SAMPLED - DEFAULT dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7649
- Clap: 0.1988
## 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: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 456
- gradient_accumulation_steps: 8
- total_train_batch_size: 8
- optimizer: Use adamw_torch with betas=(0.9,0.99) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 1.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Clap |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 9.2964 | 0.2407 | 25 | 3.6824 | 0.2393 |
| 8.452 | 0.4813 | 50 | 3.7090 | 0.2203 |
| 8.2681 | 0.7220 | 75 | 3.7505 | 0.1934 |
| 8.0758 | 0.9627 | 100 | 3.7648 | 0.2037 |
### Framework versions
- PEFT 0.13.2
- Transformers 4.47.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3 |