Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# wavlm-large_finetuned_RAVDESS
|
| 2 |
+
|
| 3 |
+
This model is a fine-tuned version of [microsoft/wavlm-large](https://huggingface.co/microsoft/wavlm-large) on the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS).
|
| 4 |
+
It achieves the following results on the evaluation set:
|
| 5 |
+
- Loss: 0.3534
|
| 6 |
+
- Accuracy: 0.9028
|
| 7 |
+
|
| 8 |
+
## Model description
|
| 9 |
+
|
| 10 |
+
More information needed
|
| 11 |
+
|
| 12 |
+
## Intended uses & limitations
|
| 13 |
+
|
| 14 |
+
More information needed
|
| 15 |
+
|
| 16 |
+
## Training and evaluation data
|
| 17 |
+
|
| 18 |
+
More information needed
|
| 19 |
+
|
| 20 |
+
## Training procedure
|
| 21 |
+
|
| 22 |
+
### Training hyperparameters
|
| 23 |
+
|
| 24 |
+
The following hyperparameters were used during training:
|
| 25 |
+
- learning_rate: 0.0003
|
| 26 |
+
- train_batch_size: 32
|
| 27 |
+
- eval_batch_size: 32
|
| 28 |
+
- seed: 42
|
| 29 |
+
- gradient_accumulation_steps: 4
|
| 30 |
+
- total_train_batch_size: 128
|
| 31 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 32 |
+
- lr_scheduler_type: linear
|
| 33 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 34 |
+
- num_epochs: 20
|
| 35 |
+
|
| 36 |
+
### Training results
|
| 37 |
+
|
| 38 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 39 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 40 |
+
| No log | 1.0 | 9 | 2.0485 | 0.2326 |
|
| 41 |
+
| 2.0712 | 2.0 | 18 | 1.8028 | 0.2917 |
|
| 42 |
+
| 1.9355 | 3.0 | 27 | 1.7300 | 0.3229 |
|
| 43 |
+
| 1.7116 | 4.0 | 36 | 1.3749 | 0.4722 |
|
| 44 |
+
| 1.4907 | 5.0 | 45 | 1.0586 | 0.6493 |
|
| 45 |
+
| 1.1558 | 6.0 | 54 | 0.8834 | 0.6771 |
|
| 46 |
+
| 0.8621 | 7.0 | 63 | 0.9206 | 0.6944 |
|
| 47 |
+
| 0.6437 | 8.0 | 72 | 0.5895 | 0.8194 |
|
| 48 |
+
| 0.4634 | 9.0 | 81 | 0.7389 | 0.7743 |
|
| 49 |
+
| 0.3974 | 10.0 | 90 | 0.4569 | 0.8542 |
|
| 50 |
+
| 0.3974 | 11.0 | 99 | 0.5140 | 0.8438 |
|
| 51 |
+
| 0.3105 | 12.0 | 108 | 0.4273 | 0.8611 |
|
| 52 |
+
| 0.2094 | 13.0 | 117 | 0.3608 | 0.8993 |
|
| 53 |
+
| 0.1401 | 14.0 | 126 | 0.5715 | 0.8194 |
|
| 54 |
+
| 0.1249 | 15.0 | 135 | 0.3715 | 0.8854 |
|
| 55 |
+
| 0.0953 | 16.0 | 144 | 0.4112 | 0.8785 |
|
| 56 |
+
| 0.0955 | 17.0 | 153 | 0.3692 | 0.9062 |
|
| 57 |
+
| 0.0807 | 18.0 | 162 | 0.4395 | 0.8646 |
|
| 58 |
+
| 0.1077 | 19.0 | 171 | 0.3413 | 0.9201 |
|
| 59 |
+
| 0.0578 | 20.0 | 180 | 0.3534 | 0.9028 |
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
### Framework versions
|
| 63 |
+
|
| 64 |
+
- Transformers 4.45.2
|
| 65 |
+
- Pytorch 2.5.0+cu118
|
| 66 |
+
- Datasets 3.0.1
|
| 67 |
+
- Tokenizers 0.20.1
|