metadata
license: apache-2.0
base_model: facebook/hubert-xlarge-ll60k
tags:
- generated_from_trainer
model-index:
- name: hubert_xlarge_emodb
results: []
hubert_xlarge_emodb
This model is a fine-tuned version of facebook/hubert-xlarge-ll60k on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8345
- Uar: 0.8889
- Acc: 0.9118
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: 0.0001
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Uar | Acc |
|---|---|---|---|---|---|
| No log | 0.2 | 5 | 1.3815 | 0.25 | 0.1985 |
| No log | 0.39 | 10 | 1.3436 | 0.5285 | 0.5956 |
| No log | 0.59 | 15 | 1.3028 | 0.5741 | 0.6618 |
| No log | 0.78 | 20 | 1.2412 | 0.6019 | 0.6838 |
| No log | 0.98 | 25 | 1.1652 | 0.75 | 0.8015 |
| 1.2216 | 1.18 | 30 | 1.0883 | 0.7315 | 0.7868 |
| 1.2216 | 1.37 | 35 | 1.0309 | 0.75 | 0.8015 |
| 1.2216 | 1.57 | 40 | 1.0217 | 0.8335 | 0.8603 |
| 1.2216 | 1.76 | 45 | 1.0084 | 0.8714 | 0.8529 |
| 1.2216 | 1.96 | 50 | 0.9415 | 0.7778 | 0.8235 |
| 0.5781 | 2.16 | 55 | 0.9293 | 0.7870 | 0.8309 |
| 0.5781 | 2.35 | 60 | 0.8470 | 0.9448 | 0.9412 |
| 0.5781 | 2.55 | 65 | 0.8673 | 0.8333 | 0.8676 |
| 0.5781 | 2.75 | 70 | 0.8454 | 0.9074 | 0.9265 |
| 0.5781 | 2.94 | 75 | 0.8139 | 0.9167 | 0.9338 |
| 0.2652 | 3.14 | 80 | 0.8254 | 0.8981 | 0.9191 |
| 0.2652 | 3.33 | 85 | 0.8233 | 0.9074 | 0.9265 |
| 0.2652 | 3.53 | 90 | 0.7989 | 0.9259 | 0.9412 |
| 0.2652 | 3.73 | 95 | 0.7939 | 0.9584 | 0.9632 |
| 0.2652 | 3.92 | 100 | 0.8093 | 0.9167 | 0.9338 |
| 0.1537 | 4.12 | 105 | 0.8138 | 0.9167 | 0.9338 |
| 0.1537 | 4.31 | 110 | 0.7898 | 0.9539 | 0.9559 |
| 0.1537 | 4.51 | 115 | 0.8138 | 0.9074 | 0.9265 |
| 0.1537 | 4.71 | 120 | 0.8463 | 0.8704 | 0.8971 |
| 0.1537 | 4.9 | 125 | 0.8643 | 0.8519 | 0.8824 |
| 0.1615 | 5.1 | 130 | 0.8137 | 0.9074 | 0.9265 |
| 0.1615 | 5.29 | 135 | 0.7750 | 0.9724 | 0.9706 |
| 0.1615 | 5.49 | 140 | 0.7745 | 0.9724 | 0.9706 |
| 0.1615 | 5.69 | 145 | 0.8123 | 0.9074 | 0.9265 |
| 0.1615 | 5.88 | 150 | 0.8693 | 0.8426 | 0.875 |
| 0.0762 | 6.08 | 155 | 0.9067 | 0.7870 | 0.8309 |
| 0.0762 | 6.27 | 160 | 0.9123 | 0.7870 | 0.8309 |
| 0.0762 | 6.47 | 165 | 0.8664 | 0.8426 | 0.875 |
| 0.0762 | 6.67 | 170 | 0.8167 | 0.9074 | 0.9265 |
| 0.0762 | 6.86 | 175 | 0.8104 | 0.9259 | 0.9412 |
| 0.1321 | 7.06 | 180 | 0.8222 | 0.8981 | 0.9191 |
| 0.1321 | 7.25 | 185 | 0.8339 | 0.8889 | 0.9118 |
| 0.1321 | 7.45 | 190 | 0.8468 | 0.8704 | 0.8971 |
| 0.1321 | 7.65 | 195 | 0.8453 | 0.8704 | 0.8971 |
| 0.1321 | 7.84 | 200 | 0.8453 | 0.8704 | 0.8971 |
| 0.027 | 8.04 | 205 | 0.8346 | 0.8889 | 0.9118 |
| 0.027 | 8.24 | 210 | 0.8292 | 0.8889 | 0.9118 |
| 0.027 | 8.43 | 215 | 0.8276 | 0.8889 | 0.9118 |
| 0.027 | 8.63 | 220 | 0.8353 | 0.8889 | 0.9118 |
| 0.027 | 8.82 | 225 | 0.8376 | 0.8889 | 0.9118 |
| 0.0499 | 9.02 | 230 | 0.8327 | 0.8889 | 0.9118 |
| 0.0499 | 9.22 | 235 | 0.8317 | 0.8889 | 0.9118 |
| 0.0499 | 9.41 | 240 | 0.8330 | 0.8889 | 0.9118 |
| 0.0499 | 9.61 | 245 | 0.8343 | 0.8889 | 0.9118 |
| 0.0499 | 9.8 | 250 | 0.8345 | 0.8889 | 0.9118 |
Framework versions
- Transformers 4.32.0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.13.3