File size: 3,733 Bytes
4d94307 ae62d1d 4d94307 ae62d1d 4d94307 ae62d1d 2f2bc0b 745031d ae62d1d 4d94307 ae62d1d 72e5e93 ae62d1d 72e5e93 4d94307 |
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 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 |
---
license: apache-2.0
base_model: facebook/wav2vec2-large-xlsr-53
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: SER_wav2vec2-large-xlsr-53_fine-tuned_1.0
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. -->
# SER_wav2vec2-large-xlsr-53_240303
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on a [Speech Emotion Recognition (en)](https://www.kaggle.com/datasets/dmitrybabko/speech-emotion-recognition-en) dataset.
This dataset includes the 4 most popular datasets in English: Crema, Ravdess, Savee, and Tess, containing a total of over 12,000 .wav audio files. Each of these four datasets includes 6 to 8 different emotional labels.
It achieves the following results on the evaluation set:
- Loss: 1.7923
- Accuracy: 0.2408
- Precision: 0.2324
- Recall: 0.2466
- F1: 0.2226
## For a better performance version, please refer to
[hughlan1214/Speech_Emotion_Recognition_wav2vec2-large-xlsr-53_240304_SER_fine-tuned1.1](https://huggingface.co/hughlan1214/Speech_Emotion_Recognition_wav2vec2-large-xlsr-53_240304_SER_fine-tuned1.1)
## Model description
The model was obtained through feature extraction using [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) and underwent several rounds of fine-tuning. It predicts the 7 types of emotions contained in speech, aiming to lay the foundation for subsequent use of human micro-expressions on the visual level and context semantics under LLMS to infer user emotions in real-time.
Although the model was trained on purely English datasets, post-release testing showed that it also performs well in predicting emotions in Chinese and French, demonstrating the powerful cross-linguistic capability of the [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) pre-trained model.
```python
emotions = ['angry', 'disgust', 'fear', 'happy', 'neutral', 'sad', 'surprise']
```
## 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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 3
- total_train_batch_size: 12
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 8
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
| 1.9297 | 1.0 | 101 | 1.9452 | 0.1233 | 0.0306 | 0.1468 | 0.0454 |
| 1.9114 | 2.0 | 202 | 1.9115 | 0.1773 | 0.1501 | 0.1803 | 0.1323 |
| 1.7863 | 3.0 | 303 | 1.8564 | 0.2081 | 0.1117 | 0.2193 | 0.1336 |
| 1.8439 | 4.0 | 404 | 1.8590 | 0.2042 | 0.2196 | 0.2156 | 0.1755 |
| 1.9361 | 5.0 | 505 | 1.8375 | 0.2081 | 0.2617 | 0.2213 | 0.1573 |
| 1.7572 | 6.0 | 606 | 1.8081 | 0.2100 | 0.2018 | 0.2214 | 0.1841 |
| 1.6715 | 7.0 | 707 | 1.8131 | 0.2389 | 0.2263 | 0.2442 | 0.2129 |
| 1.6687 | 8.0 | 808 | 1.7923 | 0.2408 | 0.2324 | 0.2466 | 0.2226 |
### Framework versions
- Transformers 4.38.1
- Pytorch 2.2.1
- Datasets 2.17.1
- Tokenizers 0.15.2
|